Language and disguistic

LANGUAGE AND LINGUISTICS

The new edition of this highly successful A–Z guide explores the

main concepts and terms used in the study of language and lin-guistics. Containing over 300 entries, thoroughly updated to reflect

the latest developments in the field, this book includes entries in:

� Cognitive linguistics

� Discourse analysis

� Phonology and phonetics

� Psycholinguistics

� Sociolinguistics� Syntax and semantics

Beginning with a brief definition, each entry is followed by a com-

prehensive explanation of the origin and usage of the term. The

book is cross-referenced throughout and includes further reading

for academics and students alike.

R.L. Trask (1944–2004) was the highly regarded Professor of Lin-guistics at the University of Sussex. His numerous publications inclu-

ded Language: The Basics (1995) and A Dictionary of Phonetics and

Phonology (1995).

Peter Stockwell is Professor of Literary Linguistics at the University

of Nottingham and the editor of the Routledge English Language

Introductions Series.

YOU MAYALSO BE INTERESTED IN THEFOLLOWING ROUTLEDGE STUDENT

REFERENCE TITLES

Language: The Basics (Second Edition)

R.L. Trask

Semiotics: The Basics

Daniel Chandler

Psycholinguistics: The Key Concepts

John Field

The Routledge Companion to Semiotics and Linguistics

Paul Cobley

LANGUAGE ANDLINGUISTICS

The Key Concepts

Second Edition

R.L. Trask

Edited by Peter Stockwell

First published as Key Concepts in Language and Linguistics 1999

This edition published 2007by Routledge

2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN

Simultaneously published in the USA and Canadaby Routledge

270 Madison Avenue, New York, NY 10016

Routledge is an imprint of the Taylor & Francis Group, an informa business

First edition text # 1999 R.L. TraskSecond edition # 2007 R.L. Trask and Peter Stockwell

All rights reserved. No part of this book may be reprinted or reproduced or utilizedin any form or by any electronic, mechanical, or other means, now known or

hereafter invented, including photocopying and recording, or in any informationstorage or retrieval system, without permission in writing from the publishers.

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication DataA catalog record for this book has been requested

ISBN 978-0-415-41358-9 (hbk)ISBN 978-0-415-41359-6 (pbk)ISBN 978-0-203-96113-1 (ebk)

This edition published in the Taylor & Francis e-Library, 2007.

“To purchase your own copy of this or any of Taylor & Francis or Routledge’scollection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”

ISBN 0-203-96113-7 Master e-book ISBN

CONTENTS

List of Key Concepts (alphabetically arranged) vi

List of Key Concepts (arranged by linguistic sub-discipline) xiPreface to the revised edition xviii

The International Phonetic Alphabet xxi

KEY CONCEPTS 1

Bibliography 330

Index 351

v

ALPHABETICAL LIST OF

KEY CONCEPTS

(Listed alphabetically, as they appear in this book)

accent

adjacency pair

adjective

adverb

affix

affricate

agreement

airstream mechanismalternation

ambiguity

analogy

anaphor

animal communication

anthropological linguistics

aphasia

apparent timeapplied linguistics

arbitrariness

argument

artificial language

aspect

autolexical grammar

autonomy

auxiliarybasic word order

behaviourism

bilingualism

bioprogram hypothesis

Black English

borrowing

case

clause

clitic

cloze

code-switching

cognitive linguistics

cohesion

colligatecommunicative competence

community of practice

comparative reconstruction

competence

complement

computational linguistics

conceptual integration

conjunctionconnotation

consonant

constituent structure

construction grammar

control

conversation analysis

conversational implicature

cooperative principlecoordinate structure

copula

creole

critical discourse analysis

critical period hypothesis

dead language

vi

deep structure

deficit hypothesis

deictic category

denotationdependency

derivation

descriptivism

design features

determiner

diachrony

dialect

diglossiadirect object

discourse

discourse analysis

displacement

distinctive feature

distribution

duality of patterning

dyslexiaelicitation techniques

ellipsis

entailment

ethics

ethnicity

ethnography of

communication

etymologyeuphemism

finite

focus

folk linguistics

foregrounding

forensic linguistics

formal grammar

framefunctionalism

functions of language

gap

gender

generative grammar

genetic hypothesis of language

genetic relationship

genre

given/new

governmentGovernment-and-Binding

Theory

grammar

grammatical category

grammatical relation

grapheme

head

historical linguisticsiconicity

identity

ideology

idiom

indicator

indirect object

Indo-European

infinitiveinflection

innateness hypothesis

integrationalism

internal reconstruction

international language

International Phonetic

Alphabet

intertextualityintonation

intuition

irrealis

kinship terms

landmark

language

language acquisition

language acquisition devicelanguage areas

language change

language contact

language death

language disability

language faculty

ALPHABETICAL LIST OF KEY CONCEPTS

vii

language family

language instinct

language myths

language planninglanguage processing

langue

lateralization

lemma

Lexical-Functional Grammar

lexicography

lexicon

lingua francalinguistic area

linguistic relativity hypothesis

linguistic sign

linguistics

literacy

loan word

localization

logonomic rulesmanner of articulation

markedness

markers

meaning

medium

mental space

mentalism

metalanguagemetaphor

minimal pair

minimalist program(me)

minority language

mirror neurons

modality

modifier

moodmorpheme

morphology

motherese

movement

name

narrative

national language

natural class

natural-language processing

neurolinguisticsneutralization

nominalization

non-verbal communication

notational convention

noun

noun phrase

number

number of languagesobserver’s paradox

official language

onomastics

open-endedness

optimality theory

oracy

origin and evolution of

languageorthography

paradigm

paradigmatic relation

paralanguage

parole

parsing

part of speech

perceptual strategyperformance

performative

person

philology

philosophy of language

phonation type

phoneme

phoneticsphonology

phonotactics

phrase

phrase-structure grammar

pidgin

place of articulation


viii

politeness

polysynthesis

possible worlds theory

powerpragmatics

predicate

preposition

prescriptivism

prestige

presupposition

productivity

pronounprosody

proto-language

proto-language hypothesis

prototype

psycholinguistics

punctuation

purism

qualitative approachquantitative approach

raising

rank scale

reconstruction

recursion

reference

relevance

rhetoricroot

rule

Saussurean paradox

schema

segment

selection restriction

semantic role

semanticssemiotics

sense

sense relation

sentence

sex differences in language

sexist language

sign language

slang

social history of language

social networksocial stratification of language

sociolinguistics

sound symbolism

speech

speech act

speech community

speech event

speech soundspeech therapy

spelling

standard language

stem

stereotype

stimulus-freedom

stress

structuralismstructure

structure-dependence

stylistics

subcategorization

subject

subordination

suprasegmental

surface structuresyllable

symbolic system

synchrony

syntactic category

syntagmatic relation

syntax

system

systematic correspondenceSystemic Linguistics

tagmemics

tense

text

text linguistics

text world theory


ix

textuality

tone language

topic

traditional grammartrajector

transcription

transformational grammar

transitivity

tree

trope

turn-taking

typologyuniversal

universal grammar

usage

utterance

variable

variation

verbverb phrase

vernacular

vocal tract

voice

voicing

vowel

well-formedness

wordword-formation

writing system

X-bar


x

LIST OF KEY CONCEPTS

(Arranged according to the linguistic sub-discipline: some items therefore

appear repeatedly)

Phonetics and phonology

accent

affricate

airstream mechanism

alternation

consonant

distinctive feature

distribution

graphemeiconicity


Alphabet

intonation

lateralization

manner of articulation

markedness

mediumminimal pair

neutralization


phonation type

phoneme

phonetics

phonology

phonotacticsplace of articulation

prosody

rule

segment

sound symbolism

speech

speech sound

stresssuprasegmental

syllable

tone language

transcription

vocal tract

voice

voicing

vowel

Morphology

affixaspect

case

copula

derivation

finite

grammar


grammatical relationgrapheme

infinitive

inflection

markedness

morpheme

morphology

paradigm

paradigmatic relationperson

polysynthesis

xi

pronoun

root

rule

spellingstem

syllable

tense

word

word-formation

Lexicology

aspect

case

clitic

colligatecopula

derivation

determiner

idiom

infinitive

inflection

irrealis

kinship termslemma

lexicography

lexicon

loan word

modality

modifier

noun

orthographypredicate

preposition

pronoun

verb

word

word-formation

Syntax

adjective

adverb

agreement

anaphor

aspect

auxiliarybasic word order

case

clause

clitic

complement

conjunction

constituent structure

controlcoordinate structure

deep structure

dependency

derivation

determiner

direct object

distribution

ellipsisfinite

frame

gap

given/new

government

grammar


grammatical relationhead

iconicity

indirect object

infinitive

intuition

markedness

modality

modifiermood

movement

nominalization


noun

noun phrase


xii

number

paradigm

paradigmatic relation

parsingpart of speech

person

phrase

phrase-structure grammar

predicate

preposition

pronoun

raisingrecursion

root

rule

sentence

structure

subcategorization

subject

subordinationsurface structure

syntactic category

syntagmatic relation

syntax

tense

topic


transitivitytree

verb

verb phrase

X-bar

Semantics and pragmatics

ambiguity

analogy

anaphor

arbitrariness

argumentaspect

clause

cohesion

colligate

connotation

conversational implicaturecopula

deictic category

denotation

entailment

euphemism

focus

frame

gendergiven/new

iconicity

idiom

irrealis

kinship terms

lemma

linguistic sign

markednessmeaning

metaphor

modality

mood

name

paralanguage

politeness

possible worlds theorypower

pragmatics

presupposition

prototype

reference

relevance

selection restriction

semantic rolesemantics

sense

sense relation

speech

stereotype

transitivity


xiii

trope

usage

utterance

verbverb phrase

Text and discourse analysis

adjacency pair

anaphor

clause

cohesion

cooperative principle

discourse

discourse analysis

ellipsisfocus

frame

functionalism

genre

intertextuality

logonomic rules

narrative

natural-language processingnon-verbal communication

oracy

politeness

possible worlds theory

rank scale

rhetoric

schema

semioticsspeech

speech act

speech community

speech event

stylistics

Systemic Linguistics

text

text linguisticstext world theory

textuality

topic

turn-taking

usage

utterancewriting system

Applied linguistics andsociolinguistics

accent

aphasia

apparent time

applied linguistics

bilingualism

bioprogram hypothesisBlack English

cloze

code-switching

colligate

community of practice

computational linguistics

conversation analysis

creolecritical discourse analysis

deficit hypothesis

dialect

diglossia

elicitation techniques

ethics

ethnicity

ethnography of communicationforensic linguistics

gender

identity

ideology

indicators

international language


Alphabetkinship terms

language change

language contact


xiv

language death

lingua franca

linguistic area

literacymarkers

minimal pair

minority language

motherese

narrative

national language

number of languages

observer’s paradoxofficial language

oracy

parole

performance

performative

pidgin

politeness

powerprestige

sex differences in language

sexist language

sign language

slang

social history of language

social network

social stratification oflanguage

sociolinguistics

speech community

speech therapy

standard language

stereotype

synchrony

tone languagetranscription

turn-taking

usage

variable

variation

vernacular

Psycholinguistics and cognitivelinguistics

bilingualism

clozecognitive linguistics

communicative competence

competence

conceptual integration

critical period hypothesis

deictic category

dyslexia

foregroundingframe

given/new

irrealis

landmark

language acquisition

language acquisition device

language areas

language disabilitylanguage faculty

language instinct

language planning

language processing

literacy

localization

mental space

mentalismmetaphor

mirror neurons

motherese

natural-language processing

neurolinguistics

oracy

perceptual strategy

possible worlds theoryprototype

psycholinguistics

qualitative approach

quantitative approach

schema

speech therapy


xv

text world theory

trajector

Historical linguistics andphilology

borrowingcomparative reconstruction

dead language

diachrony

etymology

genetic relationship

grapheme

historical linguistics

Indo-Europeaninternal reconstruction

language change

language contact

language death

language family

loan word

onomastics

origin and evolution oflanguage

orthography

philology

proto-language

proto-language hypothesis

punctuation

reconstruction

social networkstandard language

typology

writing system

Theoretical linguistics andaspects of language theory

animal communication

anthropological linguistics

arbitrariness

artificial languageautolexical grammar

autonomy

behaviourism

construction grammar

critical period hypothesisdescriptivism

design features

displacement

duality of patterning

folk linguistics

forensic linguistics

formal grammar

functionalismfunctions of language

generative grammar

genetic hypothesis of language

Government-and-Binding

Theory

innateness hypothesis

integrationalism

languagelanguage myths

langue

Lexical-Functional Grammar

linguistic relativity

linguistics

mentalism

metalanguage

minimalist program(me)natural class

open-endedness

optimality theory

parole

philosophy of language

prescriptivism

productivity

purismqualitative approach

quantitative approach

rank scale

Saussurean paradox

stimulus-freedom

structuralism


xvi

structure-dependence

surface structure

symbolic system

synchronysystem

systematic correspondence

Systemic Linguistics

tagmemics

traditional grammar


typologyuniversal

universal grammar

well-formedness


xvii

PREFACE TO THE REVISED

EDITION

The book in your hand is neither a dictionary nor an encyclo-

pedia, but something in between. As its title suggests, it pro-

vides fairly detailed coverage of over 300 key concepts in the

study of language. The named concepts selected for inclusion

are all among the most important in the field and among those

which every beginning student is likely to encounter.

The concepts are taken from every area of language study, from

traditional grammar to contemporary grammatical theory, fromchild language to language and brain, from lexicography to the

linguistic study of literary texts, from men’s and women’s speech

to language and power. Each entry provides a brief definition of

the term entered and then goes on to explain the concept in some

detail – often with numerous examples – and it also introduces

and explains related terms. Wherever possible, the historical ori-

gins of the concept are described, including the time of introduc-

tion and the names of individuals who have made the conceptprominent. When a concept is controversial, the entry says so.

These words were written by Larry Trask in 1997, when he had

completed the first edition of this Key Concepts book; seven years

later, he died after a long battle with motor-neurone disease. His

book has proven to be very popular: it is being used around the

world by the sort of new students that Trask envisaged, and it is also

regarded as a quick reference work by academics and researchers.More surprisingly, the book has found a ready audience amongst

those readers who are neither professional linguists nor full-time

students but who are simply interested in language and our recent

theories about this most definingly human characteristic.

When the publishers asked me to produce a revised edition of

this book, it was therefore with a combination of anticipation and

xviii

apprehension that I agreed. Larry Trask’s Key Concepts captures a

serious engagement with the discipline of linguistics together with

an infectious enthusiasm for language study: this produces its

characteristic style, which is authoritative, readable, sophisticated,friendly, informative and occasionally scurrilous. Updating the

entries involved removing some areas that were perhaps not as ‘key’

in the new century as in the old and adding some others that have

shown a rapid development in recent years. The rewriting has been

more interventionist and thorough than I expected, though I have

tried throughout to keep faith with Trask’s style. I have altered

only a few of his acerbic comments and have enjoyed adding some

of my own.Larry Trask trained as a chemist, and left the Allegheny moun-

tains in upstate New York for Peace Corps work in Turkey. From

there he took a degree in linguistics and a doctorate in Basque in

London, before taking up a lecturing post in Liverpool. As a stu-

dent at Liverpool University in the mid-1980s, I sat in on a few of

Larry Trask’s lectures. His desire to enlighten and enthuse was

inspiring; he was one of several colleagues there at that time who

were unwittingly responsible for the broadness of my interests inlanguage study over the intervening years. Of course he would be

unaware that the callow face in his audience twenty years ago

would be given the responsibility of adapting his book.

This revised edition, then, has been thoroughly updated and

expanded. It presents the key concepts that a newcomer to the field

is likely to come across: some of these concepts are quite complex

and advanced. I have emphasised the broadness of the title in

encompassing the slightly different concerns of ‘language’ study and‘linguistics’. There is, perhaps inevitably, a leaning towards English-

language examples, as there was in the original, though I have tried

to show where English has continuities with other languages as well

as where it is different. Unfortunately, all of the references con-

tained in the Bibliography at the end are published works in Eng-

lish, since it is the only language that I can be sure the reader of

this book understands.

The key concepts are listed alphabetically through the book. Thislist is given at the beginning, and the entries are also listed sys-

temically according to their linguistic sub-discipline. I have assumed

you will be dipping into the book here and there like a hypertext

rather than reading it like a conventional textbook, so each entry

aims to be both self-contained and intertextual. Cross-references to

other entries are given in bold face on first mention. Other terms

PREFACE TO THE REVISED EDITION

xix

which are explained in context but which do not merit their own

entry are marked in bold italics. All of these items are listed in the

Index, together with the names of key linguists mentioned in the

book. Every entry has a short list of Further Reading: I have triedto combine introductory texts with the standard and challenging

works. It should be clear from the titles which is which, but part of

the appeal of linguistic study is the shortness of the journey from

introduction to expertise, so I am happy for this book to lead the

reader into a library or catalogue and leave you to find your own

way from there.

Few people are genuinely expert in all of the areas covered by

this book. Even Larry Trask thanked Nicola Woods, Lyn Pember-ton and Richard Coates for assistance in the original composition.

In turn, I am grateful to my colleagues at the University of Not-

tingham, where the broad range of the School of English Studies

offers many chance encounters with diverse experts. Svenja

Adolphs, Ron Carter, Kathy Conklin, Zoltan Dornyei, Louise

Mullany, Norbert Schmitt, and Violeta Sotirova were witness and

assistants to the tuning up of my linguistic knowledge in the course

of compiling this book. Sara Whiteley did an excellent job of com-piling the index for me. Thanks to Joanna Gavins and Ada for

reading for clarity and demanding playtime respectively.

I am grateful to Larry Trask’s wife, Jan, for permission to revise

this Key Concepts in Language and Linguistics. I hope it appears in

a form that Larry would have approved.

Peter Stockwell

PREFACE TO THE REVISED EDITION

xx

The International Phonetic Alphabet (revised to 2005)# International Phonetic Association

xxi

LANGUAGE AND LINGUISTICS

The Key Concepts

Second Edit ion

ACCENT

A particular way in which a group of people collectively pronounce

a language. For any language with more than a handful of speakers,there are prominent regional, social and individual differences in

the way the language is pronounced by different people, and the

pronunciation of many words usually changes quite radically over

time. Sometimes these differences are dramatic. Each distinguish-

able type of pronunciation is called an accent. Depending on where

we come from and what experience we have, we will be able to

identify accents different from our own with more or less precision:

in the case of English this may be an American accent, a DeepSouth accent, a Scottish accent, a working-class London accent, a

French accent (from a non-native speaker) and so on. Speakers of

all languages can do the same.

It is important to realize that everybody has an accent: it is not

possible to speak the sounds of a language without using some

accent pattern. Of course, every one of us regards some accents as

more familiar than others, or as more prestigious than others, but

that is a different matter: we are merely more sensitive to accentswhich differ strongly from our own. In the early days of cinema, for

example, British audiences were unable to understand the American

voices in the ‘talkies’, simply because they had little experience of

hearing Americans.

In Britain, the single most prestigious accent is Received Pro-nunciation, or RP, an accent which seems to have arisen in the

prestigious ‘public schools’ (private schools) in the nineteenth cen-

tury, and was adopted as the ‘voice of the BBC’ in the 1920s. Thisaccent is not associated with any particular region, though it is

structurally most similar to certain accents of the south-east of

England. No more than 3 per cent of Britons speak with an RP

accent, though many more have a near-RP accent which differs

only in a few particulars. RP is the accent usually taught to foreign

learners of English in Britain. Nevertheless, regional and social

variation in accents within the small islands of Britain and Ireland

is very great, probably greater than anywhere else in the English-speaking world: this density is largely due to the fact that English-

speaking has its oldest history there.

In the USA, with its more recent English-speaking history, the

most distinctive accents occur down the east coast and in the south,

the areas which have been settled longest. West of the Appa-

lachians, the differences level out somewhat, with less local variation

ACCENT

3

apart from a few large cities. Regional accents of English are less

prominent in the Caribbean, still less so in Canada, and least

prominent of all in the southern-hemisphere countries such as

South Africa, Australia and New Zealand, the most recent to besettled.

The writing system of English was largely formulated in the late

medieval period with the influence of printing; since then, pro-

nunciations have changed a great deal (consider though, knight,

came, for example). Though this causes problems for foreign lear-

ners of English, the advantage is that English can be read in any

accent. Writers and journalists often use nonce-forms of spelling to

indicate specific accents of English (y’all know wadda mean?), butlinguists can use the International Phonetic Alphabet to express dif-

ferences in pronunciation very precisely.

Observe that, in the USA, an accent is usually considered to be

just one aspect of a dialect; in Britain, the two are regarded as lar-

gely independent, at least in principle.

See also: dialect; International Phonetic Alphabet; phonology

Further reading: Ball and Rahilly 1999; Chambers and Trudgill 1998;Chambers et al. 2002; Foulkes and Docherty 1999; Hughes and Trud-gill 1996; Ladefoged and Maddieson 1996; Llamas et al. 2007; Trudgilland Hannah 1994; Wells 1982.

ADJACENCY PAIR

In conversation analysis, often two utterances produced by different

interlocutors will be closely related, such as in the form question/

answer, accusation/defence, greeting/reply, farewell/reply, apology/

acceptance and so on. These two utterances are often syntactically,

semantically, lexically and cohesively linked, so they are termed

adjacency pairs. Note that adjacency pairs are not always adjacent:

they can be interrupted by intervening utterances. For example,

here is a full exchange:

1 Peter (to Nick): Can you give me a hand?

2 Peter (before Nick has a chance to answer, shouts across the

road): BOB! Any chance of a hand here?

3 Bob: Yeah, be there in a minute.

4 Nick: Well, will it take long – it’s just that I’m in a rush and—

ADJACENCY PAIR

4

5 Peter: No, a couple of seconds. Can you?

6 Nick: Sure. I’ve got to be at the station soon.

7 Peter: Thanks.

8 Nick: What do I do?9 Peter: If you could just push, I’ll be able to jump-start it.

10 Nick: OK.

There are two main adjacency pairs here: 1 and 6, and 8 and 9.

Each of these pairs has an acknowledgement follow-up (6 and 10

respectively) which is known as feedback. The first adjacency pair is

interrupted by another adjacency pair (2 and 3) which functions as

a side-sequence: it is unconnected directly with the main discourse.Nick begins an intervening question/answer adjacency pair (4 and

5) which functions as an insertion-sequence: the answer of the sur-

rounding adjacency pair (1 and 6) depends upon this one. Another

common feature of conversation appears here: Nick begins to

explain he is in a rush (4), but is interrupted (5–6) and he skip-connects back to his explanation (6). The distance that skip-

connecting is tolerated can be quite large, and skip-connects are

especially prominent in online chat-room exchanges, where otherusers intervene before an interlocutor can return to ‘their’ topic.

See also: conversation analysis; turn-taking

Further reading: Holmes 1992; Duranti 1997.

ADJECTIVE

The part of speech which includes words like big and beautiful.

English and many other languages have a large and growing class

of adjectives, though in some languages the class of adjectives is

tiny and closed or absent altogether. (In these languages, the

meanings expressed by adjectives in English are expressed by other

parts of speech.)

In English, adjectives may be identified by a number of criteria.Not every adjective exhibits every single one of the typical adjecti-

val properties, but a word that exhibits most of them must still be

classed as an adjective. Here are some tests for adjectives.

Distribution: An adjective can typically appear in each of the fol-

lowing slots to produce a good sentence: This is a(n) – book; This

ADJECTIVE

5

book is –; – though this book is, it’s not what we want. (Try this with

new, interesting, expensive, beautiful.)

Comparison: An adjective can be compared in one of the following

two ways: big/bigger/biggest; beautiful/more beautiful/most beautiful.It can also appear in the as . . . as construction: as pretty as Lisa.

Degree: An adjective can be modified by a degree modifier like very,fairly, too, so or rather: very big, fairly nice, so good, rather interesting.

Affixation: An adjective may take the prefix un- or in- to form

another adjective, the suffix -ly to form an adverb, or the suffix -ness

or -ity to form a noun, among other possibilities: happy/unhappy/

happily/happiness; possible/impossible/possibly/possibility.

Negative properties: An adjective cannot be marked for number(singular versus plural) or for tense (past versus non-past), nor can

it take the suffix -ing which goes onto verbs.

Note that these tests are grammatical in nature. Adjectives such as

big, wise, large can be adapted in deviant ways for poetic and crea-

tive effects (big him up, wise up, larging it) but in the process the

words themselves cease being adjectives and become other parts of

speech, since they are defined grammatically not semantically.The meaning of an adjective is most typically a temporary or

permanent state or condition: big, human, young, red, happy, drunk,

shiny, intelligent, asleep. Many adjectives express subjective percep-

tions, rather than objective facts: interesting, beautiful, disgusting. A

few adjectives express very unusual types of meaning: mere, utter,

the heavy of She’s a heavy smoker. In English, adjectives are placed

as pre-modifiers or post-modifiers, and the extent of this choice is

often stylistically significant: consider the differing force and pur-pose of the thick students and the students who are thick.

See also: adverb; part of speech

Further reading: Collins Cobuild 1990; Crystal 1996; Greenbaum andQuirk 1990; Hurford 1994.

ADVERB

The part of speech which includes words like soon and slowly. Eng-

lish has a fairly large class of adverbs. Most commonly, an adverbdescribes the circumstances of an action: where it is done (here,

elsewhere, overhead), when it is done (tomorrow, often, rarely, never)

ADVERB

6

or how it is done (fast, well, carefully, dramatically, resentfully). But

some adverbs have less usual kinds of meaning, and, as always, we

can only identify adverbs with confidence by their grammatical

properties.A typical property of adverbs is their position in the sentence.

Consider the sentence She poured the wine. A typical adverb like

carefully can be inserted into any one of three positions: Carefully

she poured the wine; She carefully poured the wine; She poured the

wine carefully; note that She poured carefully the wine sounds oddly

poetic, and *She poured the carefully wine is plain ungrammatical

(the asterisk marks ungrammaticality). The same is true of many

other adverbs, like often, angrily and skilfully. But not all adverbsare so flexible: yesterday and downstairs can only fit into the first

and third of the three typical positions, while fast can only fit into

the last. (Naturally, some adverbs, such as uphill, have meanings

which do not allow them to fit sensibly into this example, but con-

sider another example like She threw the ball.) Adverbs with nega-

tive or interrogative meanings do something odd when they come

first: we can’t say *Seldom she poured the wine or *Why she poured

the wine? but must say instead Seldom did she pour the wine andWhy did she pour the wine?

Two other typical properties of adverbs are their ability to be

compared (with more or most) and their ability to be modified by

words expressing degree, such as very, rather, too and so: more

carefully, most often, very skilfully, rather casually, too fast, so well.

This is usually only possible with adverbs describing how some-

thing is done, though there are a few exceptions, like often. These

adverbs can also appear in the as . . . as construction, as in Susie

drives as well as Esther.

Adverbs have few other grammatical properties. They never

change their form: for example, they cannot be marked for tense,

and they have no separate plural form (except, as always, in special

poetic deviant uses, in new dialect forms or to indicate infantile

speech: seldoming, carefuller, tomorrowed, laters).

English has a subclass of adverbs, called sentence adverbs, whichare rather different from ordinary adverbs. While ordinary adverbsdescribe some aspect of the action, the sentence adverbs express the

speaker’s view of the whole rest of the sentence. For example, in She

probably poured the wine, the sentence adverb probably says nothing

about her pouring of the wine, but rather expresses the speaker’s

view of the likely truth of the statement She poured the wine. Other

sentence adverbs are maybe, certainly, frankly, mercifully, honestly,

ADVERB

7

hopefully and fortunately. Some of these can also be used as ordin-

ary adverbs: compare the meaning of Frankly, she must tell us about

it (sentence adverb expressing the speaker’s view) with She must tell

us about it frankly (ordinary adverb describing her telling). Someprescriptivists would regard a sentence like Hopefully the ticket-

office will be open as unacceptable, since offices cannot hope, but in

fact there is almost never any ambiguity.

Observe that many adverbs describing how something is done

(the adverbs of manner) are derived from adjectives by means of the

suffix -ly: eager/eagerly, furious/furiously. But other adverbs, includ-

ing adverbs of time and adverbs of place, are usually simple words,

not derived from anything.Earlier grammarians often had the bad habit of assigning the

label ‘adverb’ to almost any troublesome word they didn’t know

what to do with, such as not, almost and very. Some dictionaries

and other books still continue this unfortunate practice today, but

in fact these words do not behave like adverbs and are not adverbs:

some of them (like very) belong to other parts of speech entirely,

while others (like not) exhibit unique behaviour and cannot be sen-

sibly assigned to any part of speech at all.

See also: adjective; part of speech

Further reading: Collins Cobuild 1990; Crystal 1996; Greenbaum andQuirk 1990; Hurford 1994.

AFFIX

A grammatical element which cannot form a word by itself. Affixes

are bound morphemes, in the sense that they are meaningful units

(morphemes) which cannot exist independently of another mor-

pheme to which they must be attached. Most (but not all) of the

world’s languages contain grammatical affixes used for various

purposes. English has fewer affixes than some other languages, but

it still has some. For example, English usually expresses plurality innouns with the sounds /s/ or /z/ as in cats and dogs, or the older -en

affix as in oxen, children, women. English of course also borrows

some plural systems from other languages in the fossilized forms

datum/data, octopus/octopi, and though we keep the -i plural in

spaghetti, we anglicize pizzas. I was once behind someone in a

queue buying ice cream ‘Magnums’ who asked for ‘three mint

AFFIX

8

Magna’! Modern linguists have debated whether the medieval

‘strong’ plural mice (inflected with an internal vowel change) should

apply to multiple examples of a computer mouse, and have decided

that mouses ought to be the right form. Other meanings expressedby affixes in English include the comparative affix -er (wider) and

the superlative affix -est (widest), the vagueness of -ish, the

abstraction of -ness, the various oppositenesses of anti-, un-, non-,

dis- and de-, the repetition of re-, the emptiness of -less, and the

anticipation of pre-(even somewhat redundantly in words like pre-

warn, pre-plan, and pre-book).

Verbs in English can also exhibit affixes, as shown by paint,

which has grammatical forms like paints (She paints pictures),painted (She painted a picture and the quite different She has pain-

ted a picture) and painting (She is painting a picture). Other affixes

can be added to the verb paint to obtain the verb repaint and the

nouns painter and painting (as in This is a nice painting).

An affix that goes on the end, like -s and -ing, is a suffix, whileone that goes on the beginning, like re-, is a prefix. Other types of

affix exist, such as infixes: observe that the Tagalog verb sulat

‘write’ has inflected forms sumulat ‘wrote’ and sinulat ‘was written’,with infixes -um- and -in- inserted into the middle of the verbal

root. English only has infixing deviantly, whether creative (abso-

bloodylutely) or poetic (pity this poor monster manunkind). There

are also superfixes, which are placed ‘on top of’ a word in the sense

of stress or pronunciation variation: note the English nouns ‘record

(a vinyl music disc) and ‘contest (a competition), distinguished from

the related verbs re‘cord and con‘test only by a change in the

placement of the stress.

See also: derivation; inflection; morpheme

Further reading: Bauer 1988; Katamba 1994.

AFFRICATE

A term relating to the manner of articulation of consonants. When

the air behind a closure is gradually released, friction of the air-

stream results. The sounds made by this sustained friction are

affricates: /s/, /z/, /S/, /Z/, /h/, /D/ and so on. Even when a plosiveis produced (/t/, /d/), there is often a very brief affrication that

follows the release: this can be sustained in certain accents such

AFFRICATE

9

as London Cockney /ts/ in Tuesday. Affricates are distinguishable in

phonetics, but cause problems in phonology, where the phonemic

value of the sounds is debatable. For example, /tS/ could be

treated contrastively in English (chip/ship, match/mats) since italters the meaning wherever it occurs in the word, and is clearly

phonologically permitted in any position. However, /pr/ cannot

occur at the end of an English word (print, leprosy) and /th/ canonly occur at the end (fifth in most accents that pronounce the

second /f/, and eighth). So should these examples be treated as

single phonological units or as separate sounds consisting of a

plosive plus an affricate?

See also: accent; phonology

Further reading: Ball and Rahilly 1999; Collins and Mees 2003; Lade-foged and Maddieson 1996.

AGREEMENT

The grammatical phenomenon in which the form of one word in a

sentence is determined by the form of another word which is

grammatically linked to it. Agreement, which is also called concord,is an exceedingly common phenomenon in languages generally, but

it is not present equally in all of them. Swahili, Russian, Latin and

German have a great deal of agreement; French and Spanish have

somewhat less; English has very little; Chinese has none at all.

Certain types of agreement are especially frequent. A finite verbmay agree in person and number with its subject. This happens in

Basque; here are the present-tense forms of joan ‘go’ (the pronouns

are in brackets since they are optional):

In each case, the form of the verb marks the subject as first,

second or third person and as singular or plural, and we say that

(ni) noa ‘I go’(hi) hoa ‘you go’ (singular intimate)Ana doa ‘Ann goes’(gu) goaz ‘we go’(zu) zoaz ‘you go’ (singular polite)(zuek) zoazte ‘you go’ (plural)Neskak doaz ‘The girls go’

AGREEMENT

10

the verb-form agrees with the subject in person and number. As you

can see from the English glosses, English has only a tiny amount of

agreement of this kind: only the third-person singular goes is expli-

citly distinguished, all other persons and numbers taking aninvariable go.

Much less frequently, a verb may agree in person and number

with its object. This also happens in Basque. The form (zuk) (ni)

ikusi nauzu ‘you saw me’ carries agreement both for the subject

‘you’ (-zu) and for the object ‘me’ (n-); compare (zuk) (gu) ikusi

gaituzu ‘you saw us’ and neskek (ni) ikusi naute ‘the girls saw me’.

Adjectives and determiners may agree in number with their head

noun. Basque does not have this, but Spanish does: compare la casa

vieja ‘the old house’ with las casas viejas ‘the old houses’, in which

both the determiner la(s) ‘the’ and the adjective vieja(s) ‘old’ show

agreement with singular casa ‘house’ and plural casas ‘houses’. As

the English glosses suggest, this kind of agreement is generally

absent from English, but we do have a trace of it in cases like this

old house versus these old houses, in which the determiner agrees

(but not the adjective).

A determiner or an adjective may also agree in case with its headnoun. This occurs in German: in mit diesem Mann(e) ‘with this

man’ (the e is optional and is therefore in brackets), the noun

Mann(e) stands in the dative case, and the determiner diesem ‘this’

agrees with it in case, while in fur diesen Mann, Mann stands in the

accusative case, and the determiner now agrees with that.

The Spanish and German examples also illustrate what might be

called agreement in gender. For example, the Spanish noun casa

‘house’ is feminine in gender; if we use instead a masculine noun,such as libro ‘book’, we get el libro viejo ‘the old book’ and los

libros viejos ‘the old books’, showing that the determiner and the

adjective are ‘agreeing’ in gender as well as in number. Such gender

matching is traditionally regarded as another variety of agreement;

strictly speaking, however, this is not agreement but government,

since a single noun like casa or libro has only one possible gender,

and hence in these cases it is not the form of the noun which

determines the forms of the other words, but its very presence – thedefining criterion for government. Some linguists apply to such

cases the label governmental concord.

See also: government

Further reading: Greville 2006; Hurford 1994.

AGREEMENT

11

AIRSTREAM MECHANISM

Any way of producing a stream of air for use in speech. We produce

speech by using our vocal organs to modify a stream of air flowingthrough some part of the vocal tract, and all speech sounds require

this airstream for their production. There are several very different

ways of producing an airstream, only some of which are used in

languages, and only one of which is used in all languages.

To begin with, an airstream may be either egressive (flowing out

of the mouth) or ingressive (flowing into the mouth). Further, the

air which is moving may be lung air (this is the pulmonic mechan-

ism), pharynx air (the glottalic mechanism) or mouth air (the vela-ric mechanism). This gives six possible combinations, only four of

which are used in speech.

In the pulmonic egressive airstream mechanism, air is squeezed

out of the lungs by the diaphragm and the rib muscles and passes

out through the mouth (and possibly the nose). This is the principal

mechanism in all languages and the only one used in most lan-

guages (including English). In the pulmonic ingressive mechanism,

air is drawn in from the outside through the mouth into the lungs;no language uses this, but you may hear it intermittently from a

child sobbing and talking at the same time.

In the glottalic egressive mechanism, the glottis is closed and the

larynx is driven up in the throat like a piston, pushing the air of the

pharynx out through the mouth. The sounds produced are ejectives,which occur in only a few languages. If the larynx is driven down-

ward instead, outside air is pulled into the mouth and pharynx, and

we have the glottalic ingressive mechanism. The sounds producedare injectives (or voiceless implosives); these are very rare in their

pure form, but, if the glottis is left open slightly, so that air can leak

out from the lungs, we get a complex ingressive-egressive mechan-

ism, producing voiced implosives, which are much commoner.

In the velaric egressive mechanism, the back of the tongue is

pressed against the velum and another closure is made in front of

this; the tongue body is pushed up, so that, when the front closure

is released, mouth air is driven outward. The resulting sounds arereverse clicks, which do not occur in any language. If, instead, the

tongue body is pulled downward, when the front closure is released,

air is pulled into the mouth; this is the velaric ingressive mechanism,

and the resulting sounds are clicks. Clicks occur as speech sounds

in some languages of southern Africa; elsewhere, these sounds

occur only paralinguistically, as in the English tsk-tsk noise of

AIRSTREAM MECHANISM

12

disapproval or the clack-clack noise for geeing up a horse. All

speech sounds can be described in these ways by their manner of

articulation, even non-linguistic sounds: a velaric ingressive bilabial

plosive is a kiss.There is one other airstream mechanism, which is very unusual.

Persons who have had their larynxes removed surgically can learn

to produce an airstream by swallowing air and then forcing it up

through the oesophagus; this oesophagic egressive airstream is

effectively a controlled belch.

See also: phonation type

Further reading: Ball and Rahilly 1999; Collins and Mees 2003; Crystal1997; Ladefoged and Maddieson 1996; Laver 1994.

ALTERNATION

A variation in the form of a linguistic element depending on where

it occurs. Certain English nouns ending in the consonant /f/ formtheir plurals with /v/ instead: leaf but leaves, knife but knives. We

say that such items exhibit an /f/–/v/ alternation. For most (not all)

speakers a similar alternation occurs in singular house (with /s/) but

plural houses (with /z/), though here our spelling system does not

represent the alternation explicitly.

A somewhat different sort of alternation is found in related

words like electric (which ends in /k/) and electricity (which has /s/

instead of /k/ in the same position).More subtle is the three-way alternation occurring in the English

plural marker. The noun cat has plural cats, pronounced with /s/,

but dog has plural dogs, pronounced with /z/ (though again the

spelling fails to show this), and fox has plural foxes, with /z/ pre-

ceded by an extra vowel. This alternation is regular and predictable;

the choice among the three alternants (as they are called) is deter-

mined by the nature of the preceding sound.

Alternations are exceedingly common in the world’s languages,and they are often of great interest to linguists trying to produce

elegant descriptions of languages. Where pronunciation changes at

a grammatical boundary, as in the examples at morpheme bound-

aries above, these are called sandhi, a term deriving from the

ancient Sanskrit grammarians. Changes to morpheme pronuncia-

tion within a single word are internal sandhi; external sandhi also

ALTERNATION

13

occurs across word boundaries, as in the introduction of an ‘intru-

sive /r/’ between ‘law and order’, a /w/ between ‘mellow elephant’,

or the merging of word-final /t/ and word-initial /j/ in ‘don’t you’

into a single affricate /tS/.

See also: phonetics; phonology; phonotactics

Futher reading: Ball and Rahilly 1999; Bloomfield 1933; Hockett 1958;Lass 1984; Matthews 1991; Sommerstein 1977.

AMBIGUITY

Two or more sharply distinct meanings for a single string of words.

The simplest type of ambiguity is a lexical ambiguity, which results

merely from the existence of two different meanings for a single

word. Example: The sailors enjoyed the port. Here port can mean

either ‘fortified wine’ or ‘town by the sea’, and the entire string of

words accordingly has two different interpretations, but the struc-

ture of the sentence is exactly the same in both cases. Where inten-tional, such ambiguity often has humorous, ironic or subversive

effects.

More interesting are structural ambiguities, in which the words

have the same meanings, but quite different structures can be

assigned to the entire string of words, producing different meanings.

Examples: Small boys and girls are easily frightened; Exploding

mines can be dangerous; The shooting of the hunters was appalling;

Anne likes horses more than Mark. In the first two of these, thedifferent structures can be easily represented by tree diagrams, and

such cases are called surface-structure ambiguities. In the last two,

the tree structures appear to be identical in both readings (inter-

pretations), and we need to appeal to more abstract levels of

representation to identify the differences in structure; these are

deep-structure ambiguities.Complex cases are possible, involving both lexical and structural

ambiguities, as in the classic Janet made the robot fast, which has anastonishing number of quite different readings.

The concept of ambiguity can be extended to cases which are

only ambiguous when spoken, and not when written. Simple cases

of this are an ice-box versus a nice box or a slide-rule versus a sly

drool. The mishearing of song lyrics demonstrates the phenomenon,

where they are known as mondegreens after a Harper’s magazine

AMBIGUITY

14

columnist in 1954 confessed to mishearing a folk-ballad lyric They

ha’e slain the Earl of Murray, and they laid him on the Green as They

ha’e slain the Earl of Murray and Lady Mondegreen. Other exam-

ples include: all eyes and chest for Simon and Garfunkel’s all lies

and jest; The ants are my friends, they’re blowin’ in the wind for

Dylan’s The answer, my friend, is blowin’ in the wind; I’m a turtle

crawlin’ out of Michelle for Shaggy’s I’m a turtle crawlin’ outta me

shell; and In the meadow we can build a snowman; then pretend that

he is sparse and brown for the actual In the meadow we can build a

snowman; then pretend that he is Parson Brown, from ‘Winter

Wonderland’.

See also: meaning; structure

Further reading: Cruse 1986; Kempson 1977.

ANALOGY

A type of language change in which some forms are deliberatelychanged merely to make them look more like other forms. The

ordinary processes of language change, including perfectly regular

changes in pronunciation, can have the effect of introducing irre-

gularities. Speakers sometimes react to the presence of irregularities

in their language by eliminating them and making the irregular

forms regular; this is one kind of analogy.For example, when Latin was changing into French, the pro-

nunciation of stressed /a/ and that of unstressed /a/ developed dif-ferently, in a perfectly regular manner: stressed /a/ became the

diphthong /ai/, while unstressed /a/ remained /a/. This led to appar-

ently irregular variations in the stems of certain verbs, as with the

verb meaning ‘love’; compare the first two columns in the following

table (here an acute accent marks the position of the Latin stress):

Latin Old French Mod. French

1Sg amo aim aime2Sg amas aimes aimes3Sg amat aimet aime1Pl amamus amons aimons2Pl amatis amez aimez3Pl amant aiment aiment

ANALOGY

15

As you can see, the stem of the verb fluctuated between aim- and

am- in Old French in a seemingly unpredictable way (the Latin

stress was also lost in Old French). As a result, speakers subjected

the forms in am- to analogy, producing the modern French formsshown in the third column (there have been further changes in

pronunciation, of course, but these are not relevant here).

Analogy can be far less systematic than this, and it can even turn

regular forms into irregular ones. Many formerly irregular English

verbs have been turned into regular verbs by the analogy of cases

like love/loved; an example is work/wrought, which has been analo-

gized to work/worked. On the other hand, in Early Modern English,

the past tense of catch was the regular catched, but this has beenreplaced by caught, apparently by analogy with taught, and many

Americans have replaced dive/dived with dive/dove, by analogy with

verbs like drive/drove.

See also: language change; semantics

Further reading: Trask 1996.

ANAPHOR

A linguistic item which takes its interpretation from something else

referred to in the same sentence or discourse. In the sentence Susie

wants to get a job in Paris, but she needs to improve her French first,

the item she, in the most obvious interpretation, means Susie. We

say that she is an anaphor, and that Susie is the antecedent of she;the relationship between these two items is one of anaphora, or

binding, and she is bound by Susie. Further examples of anaphors

include herself in Susie injured herself (antecedent Susie) and each

other in Susie and Mike are seeing a lot of each other (antecedent

Susie and Mike). Anaphoric reference involves ‘pointing back’ to

the antecedent, where the antecedent is often the most fully realized

lexical item. Sometimes the less full item occurs first (She needs to

improve her French, Susie does, or He knew something was wrong

when he arrived. Henry looked around himself anxiously). This phe-

nomenon of ‘pointing forward’ is called cataphora. Likewise, theantecedent of an anaphor need not be in the same sentence. Con-

sider this: Susie is looking run-down. I think she needs a holiday.

Here the antecedent Susie is in a different sentence from the anaphor

she which points to it.

ANAPHOR

16

It is possible for a zero-element (a null element) to be an ana-

phor: instead of saying Susie needs a new car but she doesn’t have

the money, we can say Susie needs a new car but doesn’t have the

money. In the second version, instead of the overt anaphor she, wehave only a piece of silence, but the interpretation is the same. For

linguistic purposes, we often write the second version with the

symbol ø (meaning ‘zero’) or e (for ‘empty’) in the appropriate

place: Susie needs a new car but e doesn’t have the money. The zero

anaphor represented as e is often called an empty category.In linguistic descriptions, it is common practice to use referential

indices, usually subscript letters, to indicate explicitly which anaphors

have which antecedents; items which are coindexed (have the samesubscripts) are coreferential (refer to the same thing), while those which

have different subscripts refer to different things. So, for example,

Mikei has found hisi dog means ‘Mike has found his own dog’, while

Mikei has found hisj dog means ‘Mike has found somebody else’s dog’

(here the preceding context must make it clear who owns the dog).

Anaphora in general, and empty categories in particular, pose

many intricate problems of linguistic analysis, and in recent years

they have been the object of intensive investigations. Theoreticallinguists are fascinated by the seemingly complex nature of the rules

governing the use of anaphors, and grammatical theorists have

often seen the elucidation of these rules as a matter of fundamental

importance. At the same time, functional linguists are deeply inter-

ested in the ways in which anaphors are used to structure dis-

courses. Psycholinguists are interested in whether anaphoric reference

is to the pre-text or to the situation modelled mentally out of that

text, and linguists with typological or anthropological interests havedevoted considerable attention to the various ways in which ana-

phors are employed in different languages.

See also: gap; pronoun

Further reading: Barss 2003; Huddleston 1984.

ANIMAL COMMUNICATION

The signalling systems used by non-human creatures. Most of the

other creatures on the planet can communicate with other members

of their species in one way or another, and often by specialized

vocal noises termed calls. But the signalling systems of these creatures

ANIMAL COMMUNICATION

17

are vastly different from human language. First, they lack duality of

patterning: they are based on the principle of ‘one sound, one

meaning’, and neither sounds nor meanings can be modified or

combined. Consequently, they lack open-endedness: only a tiny numberof different meanings can be expressed. They lack displacement:

‘utterances’ are confined to the here and now. They lack stimulus–

freedom: a call is produced always and only when the appropriate

stimulus is present, and there is no choice.

Hence non-human creatures live in a communicative world which

is alien to us: it is bounded by the horizon, lacking a past or a

future, consisting only of the endless repetition of a few familiar

messages about what’s going on at the moment. Moreover, unlikehuman languages, with their ceaseless and rapid changes, the sig-

nals used by other species never change by any process faster than

evolutionary change.

There are marginal exceptions: honeybee dances contain a lim-

ited amount of displacement; bird songs possibly contain an ele-

ment of duality; whale songs change from year to year; a fox may

occasionally give a danger call in the absence of any danger merely

to distract her cubs from a meal she is trying to eat. But theseexceptions are inconsequential: animal signals do not remotely

approximate to human language, and they cannot be regarded as

simpler versions of it.

Since all these statements are true of our closest living relatives,

the apes, it follows that our non-human ancestors of a few million

years ago also had such a limited system. Attempts at teaching

scaled-down versions of human language to apes and other crea-

tures have often been vitiated by poor procedure, but there is now asmall amount of evidence suggesting that these creatures, when

intensively trained under laboratory conditions, can learn at least

the rudiments of a human language, though no more.

See also: design features; origin and evolution of language; protolanguagehypothesis

Further reading: Crystal 1997; Malmkjær 2004; Steinberg 1993; Trask1995; Wallman 1992.

ANTHROPOLOGICAL LINGUISTICS

The study of the relation between language and culture. Anthro-

pologists generally find it necessary to learn the languages of the

ANTHROPOLOGICAL LINGUISTICS

18

people they are studying, and they realized early that the languages

themselves might provide valuable clues about the cultures under

investigation. In the late nineteenth century, the anthropologist

Franz Boas in the USA laid particular stress upon the importanceof Native American languages in the study of Native American

cultures and, thanks to his influence and that of his student Edward

Sapir, American linguistics was largely born out of anthropology.

As a result, American linguistics long retained an anthropological

orientation, and indeed most linguists worked in anthropology

departments until the middle of the twentieth century. In Britain,

the anthropologist Bronislaw Malinowski drew comparable atten-

tion to the study of languages in the early twentieth century, andhis ideas greatly influenced J.R. Firth, widely regarded as the foun-

der of linguistics in Britain.

The explosive growth of areas like sociolinguistics, psycholinguis-

tics and linguistic theory during the past few decades has not

obliterated interest in anthropological linguistics, which is now

recognized as a distinctive discipline in its own right; it is sometimes

also called linguistic anthropology. Anthropological linguists have

been greatly interested in such topics as kinship systems, colourterms, metaphors, systems for conferring names upon people and

places, connections between languages and myths, folk taxonomies(systems for classifying animals and plants), the treatment of space

and time in languages, the expression of sex differences and social

differences in speech, and the structure of narratives; more than

most linguists, they have been intrigued by the linguistic relativity

hypothesis.

See also: cognitive linguistics; ethnography of communication; linguisticrelativity

Further reading: Bonvillain 1993; Crystal 1997; Duranti 1997; Foley1997; Palmer 1996.

APHASIA

Disordered language resulting from brain damage. Strictly speak-

ing, we should say dysphasia (which means ‘disordered speech’) for

damage to the language faculties and reserve aphasia (which means

‘absence of speech’) for cases in which the victim’s language facul-

ties are totally destroyed. However, in practice the two terms are

APHASIA

19

used interchangeably, and the most profound and severe cases are

distinguished as global aphasia.That damage to the head can produce language disorders

has been known since ancient times. But it was only in the mid-nineteenth century that scientists began to investigate the pro-

blem systematically, by studying the symptoms of brain-damaged

patients and then, after the victims’ deaths, by carrying out post-

mortem examinations to see which areas of the brain had been

damaged. The French surgeon Paul Broca found that damage to a

particular area of the brain, with a high degree of consistency,

produced an aphasia characterized by painful, halting speech and

a near-total absence of grammar; today the area he identified iscalled Broca’s area, and the associated aphasia is called Broca’saphasia. A few years later, the German neurologist Carl Wernicke

identified a second area of the brain, damage to which consistently

produces a different aphasia, characterized by fluent but senseless

speech and grave difficulties in comprehension: we now speak of

Wernicke’s area and Wernicke’s aphasia. Broca’s area and Wer-

nicke’s area are now known to be two of the most important lan-

guage areas in the brain, each with responsibilities for specificaspects of language.

In practice, no victim ever suffers damage exclusively to one

neatly defined area of the brain, and consequently every sufferer

exhibits a somewhat distinctive range of symptoms. The complexity

of aphasia rather discouraged further research during much of

the twentieth century. However, in the middle of that century the

American neurologist Norman Geschwind revived interest in

the subject, confirmed the existence of distinct language areas in thebrain, and developed a classification of aphasias which is now more

or less standard. For example, Broca’s and Wernicke’s areas are

connected by a bundle of fibres called the arcuate fasciculus, anddamage to this produces a third type of aphasia, conduction apha-sia, with specific symptoms of its own. Recent evidence from mag-netic resonance imaging (MRI scans) of the brain is confirming

many of the hypotheses of early neurolinguistics.

It has also been discovered that brain damage affects users ofsign language in precisely the same way as it affects users of spoken

language. This confirms that what is damaged is language, and not

merely the ability to speak or to perceive speech. It is crucial to

distinguish aphasia from speech defects, which result merely from

damage to the nerves or muscles controlling the speech organs, and

have no consequences for the language faculty itself.

APHASIA

20

See also: language areas; language disability; neurolinguistics; speechtherapy

Further reading: Crystal 1997; Malmkjær 2004; O’Grady et al. 1996.

APPARENT TIME

A technique for studying language change in progress. One way of

studying language change in a community is to examine the speech

of that community at intervals over several generations. But such

real-time studies are often not practical: several generations ofresearchers might be involved; universities are unlikely to fund such

long-term projects; and techniques of fieldwork and analysis will

change so radically that early data become unreliable.

An alternative is the use of apparent time. In this approach, we

begin by assuming that individuals normally acquire their speech

habits early in life and thereafter rarely change them. If this is

so, then a comparison of the speech of elderly, middle-aged

and younger speakers in a community will reveal any linguisticchanges which are in progress: the younger the speakers, the more

conspicuously their speech will be affected by any changes in

progress.

Pioneered by the American linguist William Labov in the 1960s,

apparent-time studies have proved to be a powerful tool in exam-

ining language change, even though it has been found that, in cer-

tain circumstances, older speakers actually do change their speech

later in life.

See also: language change; quantitative approach; sociolinguistics

Further reading: Labov 1994; Llamas et al. 2007.

APPLIED LINGUISTICS

The application of the concepts and methods of linguistics to any of

various practical problems involving language. The term appliedlinguistics is most often encountered in connection with foreign-

language teaching. Linguistic concepts find a number of uses here,

for example in contrastive linguistics – the systematic comparison of

the sounds, the words and the grammatical systems of the mother

APPLIED LINGUISTICS

21

tongue and the language being learned – with the intention of

bringing out the important differences which need to be mastered.

But linguistics has also proved useful in a variety of other prac-

tical domains, such as mother-tongue teaching, lexicography, trans-lation, the teaching of reading, forensic linguistics, and the diagnosis

and treatment of language disability. Today all these are understood

as forming part of applied linguistics, such that the term is used by

many simply in contrast with theoretical linguistics to emphasize its

practical, fieldwork-based, data-driven and empirical nature.

See also: language planning; sociolinguistics; stylistics

Further reading: Cook 2003; Corder 1975; Cotterill 2002; Crystal 1997;Davies 1999; Emmitt et al. 2006; O’Grady et al. 1996; Pennycook 2001;Richards et al. 1992; Schmitt 2002.

ARBITRARINESS

The absence of any necessary connection between the form of aword and its meaning. Every language typically has a distinct word

to denote every object, activity and concept its speakers want to

talk about. Each such word must be formed in a valid manner

according to the phonology of the language. But, in most cases,

there is absolutely no reason why a given meaning should be deno-

ted by one sequence of sounds rather than another. In practice, the

particular sequence of sounds selected in a given language is

completely arbitrary: anything will do, so long as speakers agreeabout it.

Speakers of different languages, of course, make different choices.

A certain large snouted animal is called a pig in English, a Schwein

in German, a cochon in French, a cerdo in Spanish, a mochyn in

Welsh, a txerri in Basque, a numbran in Yimas (a language of New

Guinea), and so on across the world. None of these names is more

suitable than any other: each works fine as long as speakers are in

agreement.Such agreement need not be for all time. The animal was for-

merly called a swine in English, but this word has dropped out of

use as a name for the animal and been replaced by pig.

Arbitrariness can be demonstrated the other way round. Many

languages allow a word to have the phonetic form [min], but there

is no earthly way of predicting the meaning of this word if it should

ARBITRARINESS

22

exist. In English, [min] (spelled mean) exists and has several unre-

lated meanings: ‘tight-fisted’, ‘cruel’, ‘average’, ‘signify’. French

mine means ‘(coal) mine’; Welsh min is ‘edge’; Irish min is ‘meal’;

Basque min is ‘pain’; Arabic min is ‘from’. There is nothing aboutthis sequence of sounds that makes one meaning more likely than

another. Arbitrariness is pervasive in human languages (and also in

animal communication), but there does nonetheless exist a certain

amount of iconicity: cases in which the relation between form and

meaning is not totally arbitrary.

Unfortunately, even with some iconicity, it is the presence of

massive arbitrariness which makes impossible the universal transla-tor beloved of science-fiction films, unless the machine worked bytelepathy rather than linguistics. Because of arbitrariness, even the

most powerful computer program can have no way of guessing the

meaning of a word it has not encountered before.

Linguists have long realized the importance of arbitrariness, but

it was particularly stressed by the Swiss linguist Ferdinand de Saus-

sure in the early twentieth century, with his concept of the linguistic

sign.

See also: design features; iconicity; linguistic sign; semantics; soundsymbolism


ARGUMENT

Any one of the noun phrases required by a particular verb. Each

verb requires some number of noun phrases (NPs) to accompany it

in a sentence, if the result is to be grammatical; these NPs are its

arguments, and the number of NPs required by a verb is the valencyof that verb. For example, the simple intransitive verbs smile and

arrive are monovalent, requiring only one argument, the subject, as

in Susie smiled. (Note that *Susie smiled Natalie is ungrammatical;

the asterisk marks this). In contrast, a simple transitive verb likekiss or slap is divalent, requiring a subject and one object, as in

Susie kissed Natalie. (Note that *Susie kissed is ungrammatical).

But a ditransitive verb like give or show is trivalent, requiring a

subject and two objects, as in Susie gave Mike a present. A given

English verb usually requires between one and three arguments, but

note the unusual behaviour of the verb rain, which neither requires

ARGUMENT

23

nor permits any arguments at all, except for the ‘dummy’ subject it,

as in It’s raining. This verb arguably has a valency of zero.

In addition to its arguments, a verb very often permits some further

phrases which are optional. These optional phrases are adjuncts,and adjuncts in English are most often expressed as prepositionalphrases or as adverbial phrases. For example, the minimal sentence

Susie kissed Natalie can be expanded with some optional adjuncts

to yield Susie kissed Natalie on the neck in the kitchen this morning

(there are three adjuncts here).

See also: complement; grammatical relation; transitivity

Further reading: Butt and King 2000; Kroeger 2004; van Valin 2001.

ARTIFICIAL LANGUAGE

A language deliberately invented by a particular person. Since the

seventeenth century, hundreds of artificial languages have been

invented; some of them were no more than sketches, while otherswere provided with extensive grammars and large vocabularies.

Only a handful of these have ever gained any use.

Many of the earlier attempts were made by philosophers, and

these were often a priori in nature, meaning that they paid no

attention to existing languages but were put together according to

whatever principles seemed good to the inventors. More than a few

were intended to be ‘universal’ or ‘logical’ languages and were

based on some grand scheme for classifying all human knowledge.All were deeply impractical. Among the more notable attempts

were those of the Frenchman Descartes, the Scot Dalgarno and the

Englishman Wilkins.

Since the nineteenth century, artificial languages have more

usually been a posteriori, that is, derived in some way from existing

languages, and they have been constructed by linguists, logicians,

priests, politicians, oculists and businessmen. The German philoso-

pher Leibniz had much earlier proposed a kind of regularized Latinwith only a small number of endings, and the French politician

Faiguet had likewise sketched out a kind of regularized French, in

both cases to no effect. In 1880 the German priest Schleyer pub-

lished Volapuk, an enormously awkward and complex mixture of

bits of several European languages with cumbersome grammatical

endings of his own devising; the result resembled a kind of demented

ARTIFICIAL LANGUAGE

24

Swedish, but it attracted hundreds of thousands of followers for a

few years. Then in 1887 the Polish oculist Zamenhof published

Esperanto, a much simpler language also stitched together from

bits and pieces of several European languages, and this has eversince been the world’s single most widely learned and used artificial

language.

Esperanto still has a number of cumbersome features, and sim-

plified versions called Ido, Esperantido, Espido, Esperantuisho and

Modern Esperanto have been constructed, with minimal success.

The Danish linguist Jespersen constructed a greatly modified off-

shoot called Novial, which attracted little interest. The Italian logi-

cian Peano invented Latino Sine Flexione, a kind of simplified andregularized Latin, and the American writer Hogben followed with

Interglossa, essentially a version of classical Greek with no word-

endings at all. A kind of regularized common Romance called

Interlingua was constructed by a group of people and has found

some limited use, especially for scientific purposes. A modified and

expanded version of Interglossa now called Glosa has recently been

published. And dozens of other twentieth-century projects have

been put forward only to sink without trace.Worthy of mention is Basic English, a stripped-down version of

English employing only 850 words and published in 1930 by the

British scholar C.K. Ogden; this was popular for a while, but it

proved unworkable and has disappeared.

All of these conlangs (constructed languages) are of the auxlang(auxiliary language) type. However, there are many artlangs devel-

oped for artistic purposes. The fantasy writer J.R.R. Tolkien

invented not one but several languages for his novel The Lord of the

Rings, and he even provided some of them with histories, sound

changes, family trees and writing systems (most of which draw on

Finnish, Old Icelandic and Welsh). The linguist Marc Okrand was

commissioned to invent a Klingon language for the Star Trek films,

and there are now Klingon translations of the Bible and Shake-

speare. Lastly, loglangs (logical languages) such as computer-

programming languages can have a practical purpose, though most

(like ‘Loglan’ and its descendant ‘Lojban’ maintained by the Logi-cal Language Group) display a fundamental wrong-headedness

about what a language is and what it is used for.

See also: language; natural-language processing

Further reading: Bodmer 1944; Crystal 1997; Large 1985; Malmkjær 2004.

ARTIFICIAL LANGUAGE

25

ASPECT

The grammatical category representing distinctions in the temporal

structure of an event. Quite independently of its location in time, anevent may be viewed as having any of a number of different tem-

poral organizations: it may be seen as having internal structure or

as consisting of an unanalysable whole; it may be seen as extending

over a period of time or as occurring in a single moment; it may be

seen as a single occurrence or as a series of repeated occurrences; it

may be seen as beginning, continuing or ending. All these and

others are types of aspect.Compared with many other languages, English has a rather modest

aspectual system, but aspect is still important in English. Here are

some examples. The sentence She smoked illustrates perfective aspect(the event is viewed as an unanalysable whole); She was smoking shows

continuous (or progressive) aspect (the event is viewed as extending

over time); She used to smoke exhibits habitual aspect (the event is

viewed as a customary or habitual one); She kept smoking illus-

trates iterative aspect (the action is viewed as a series of repeated

events); She started smoking exhibits inchoative aspect (the event isviewed as just beginning); and She quit smoking exhibits conclusiveaspect (the event is viewed as drawing to a close). The second, third

and fourth of these all represent differing types of imperfective aspect(the action is viewed as having some kind of internal structure).

Some other languages display further aspectual forms, such as

the punctual aspect (the event is viewed as occurring in a single

moment); English has no special form for this, and we use our

perfective form, as in She sneezed. But compare Basque, whichdistinguishes Agertu zen ‘She appeared’ (for a moment) from

Agertzen zen ‘She appeared’ (over a period of time). The African-

American variety of English has a completive aspect (She done talked)and Jamaican English has a narrative aspect (She bin talked).

English has another distinctive form, the perfect, which has sev-

eral functions but most typically expresses a state resulting from an

earlier event. For example, the perfect form She had finished the

wine most obviously means ‘There was no wine then because shedrank the last of it earlier’. The perfect is often classed as an

aspect, although it is decidedly unusual among aspects. (Note that

perfective and perfect have very different meanings; even some

textbooks get these two confused).

Aspect must be carefully distinguished from tense, even though

the formal expression of the two categories is often deeply intertwined

ASPECT

26

in languages. All of the English examples above are in the past

tense, and all of them have corresponding non-past (‘present’)

forms, except that the perfective She smoked and the habitual She

used to smoke both have the same present-tense counterpart, Shesmokes, which most often has the habitual sense. Crudely, English

has only two tenses (past and present) marked morphologically; all

other time characteristics are expressed by aspect.

Certain types of aspect-like distinctions may be expressed by

lexical means, rather than grammatically; an example is the con-

trast among English nibble, eat, devour. A distinction expressed in

this way is called an Aktionsart or lexical aspect. Similarly, a verb

can carry a sense of its own duration: to make a table carries itsown completion in a way that to sit at a table does not (it would be

odd to talk of finishing sitting): a verb like the former with a natural

ending is telic, and the latter verb is atelic.

See also: grammatical category; tense

Further reading: Binnick 1991; Comrie 1976; Dahl 1985.

AUTOLEXICAL GRAMMAR

A model of grammar proposed by Jerrold Sadock in the 1980s, inwhich morphology, syntax and semantics are regarded as indepen-

dent modules with a phrase-structure grammar. The modules are

brought together at the interface by the lexicon, which matches up

the three modules. Only sentences which do not violate the rules of

all three modules are regarded as well-formed sentences. Autolexical

grammar was developed from the viewpoint of languages which

feature morphosyntactic mismatches (such as the Kalaallisut of

Western Greenland), and then subsequently applied to English,where it accounts particularly well for spoken forms.

See also: lexicon; phrase-structure grammar

Further reading: Sadock 1991.

AUTONOMY

The view that the human language faculty is independent of general

mental and cognitive abilities. A young child is obliged to spend

AUTONOMY

27

years learning to make sense of the world it is born into, and at the

same time it must learn its first language. For decades there has

been a controversy, sometimes called the nature-nurture debate or

the content-process debate, over whether children are born with adistinctive and largely independent faculty for learning language, or

whether they simply acquire a language in the same way they

acquire other kinds of understanding and skills, by using their

general all-purpose cognitive abilities.

The first view – represented by such proposals as the genetic

hypothesis of language and the innateness hypothesis – is probably

supported by a majority of theoretical linguists. The second has

been supported by a number of psychologists, and has recently beenmost strongly articulated within cognitive linguistics. The first view

holds that children are born with specialized structures or areas in

their brains which are dedicated to the learning and use of lan-

guages; Chomsky’s version further holds that important informa-

tion about the nature of human languages is already present at

birth. The second view denies this, and sees language acquisition as

not different in kind from, say, learning to judge size and distance;

some versions go further and claim that learning a first language isnot different from learning to ice skate or to drive a car.

This last, extreme, view can probably be disposed of: the abun-

dant evidence for the critical period hypothesis, demonstrating that

first-language acquisition is rapid in children but impossible in

adults, surely demonstrates that learning a first language is very

different from learning to ice skate. Otherwise, though, the debate is

still very much alive.

Linguists like to support the first view by pointing to the evi-dence from language disability: some disabilities, such as the Wil-liams syndrome, appear to leave the language faculties intact while

severely damaging other mental faculties; others, such as SpecificLanguage Impairment, chiefly affect only linguistic behaviour while

leaving other mental faculties largely unscathed. Supporters of the

second view point to continuities and common properties between

language and other visual, aural, tactile and spatial development.

See also: cognitive linguistics; critical period hypothesis; genetic hypoth-esis of language; innateness hypothesis; language disability; languagefaculty

Further reading: Bates 1976; Bates et al. 1979, 1988; Jackendoff 1993;Macwhinney and Bates 1989; Pinker 1994.

AUTONOMY

28

AUXILIARY

A specialized grammatical item, most often part of a verb phrase,

which serves to express any of several grammatical categories. TheEnglish auxiliaries are specialized verbs; they chiefly serve to express

aspect, voice and modality, and they commonly also carry markers

of tense and agreement.

The English auxiliary verbs possess several properties not shared

by ordinary verbs.

� Negation: She is running; She isn’t running/She is not running.

Compare She smokes; *She smokesn’t; *She smokes not (theasterisk marks ungrammaticality).

� Inversion: She is running; Is she running?

Compare She runs; *Runs she?

� Ellipsis: Susie is running, but Janet isn’t.

� Emphasis: She did run.

� Tagging: She did run, didn’t she?

The English auxiliaries are of two types. The primary (or non-modal) auxiliaries are be and have. These resemble ordinary main

verbs in having a full set of inflected forms (have, has, had, having),

and a primary auxiliary can be the only verb in a sentence: She is

British; He has brown eyes. The modal auxiliaries have only one or

two forms each; most of them come in pairs, which can to some

extent be regarded as differing in tense: can/could, will/would; shall/

should; may/might; must. These lack the -s ending of the third-

singular present: *She cans speak French.The verbs need, dare and ought sometimes exhibit some (but not

all) of the properties of modal auxiliaries; these are called the semi-modals. Examples: Need she come? She dare not do it, but she ought

to. The verb do serves as a ‘dummy’ auxiliary, inserted to carry the

auxiliary properties when no other auxiliary is present. Examples:

She doesn’t smoke; Does she smoke?; Janet doesn’t smoke, but Susie

does; She does smoke.

In some other languages, such as Australian languages, the aux-iliaries are not verbs at all, but an entirely distinct part of speech.

Some linguists prefer to take the same view of the English auxiliaries.

See also: subcategorization; verb; verb phrase

Further reading: Coates 1983; Frawley 2006; Hurford 1994; Palmer 2001.

AUXILIARY

29

BASIC WORD ORDER

The most typical order of elements in the sentences of a language.

Almost every language shows a strong preference to put the wordsof a sentence into a particular order; this preferred order may be

virtually rigid – with almost no departures allowed – or it may be

little more than a statistical preference. This preference is the basicword order of the language.

A convenient and widely used way of characterizing basic word

order is in terms of just three major elements: Subject, Object andVerb, or S, O and V. The basic word order of English, in these

terms, is SVO: we normally say The Turks love backgammon, whileBackgammon the Turks love is unusual, and other orders, such as

*Love the Turks backgammon or *Backgammon love the Turks are

impossible. SVO order is also typical of French, Swahili and

Chinese, among others.

Other languages have different basic word orders. VSO is found

in Irish and Welsh, SOV in Japanese, Turkish, Basque and Que-

chua, and VOS in Malagasy (in Madagascar). The Amazonian

language Hixkaryana is OVS, and there are reports that anotherAmazonian language, Apurina, may be OSV.

It appears that SOV order is the most frequent on the planet,

followed closely by SVO and more distantly by VSO. VOS is

decidedly uncommon, and OVS and OSV are, at best, very rare. No

one knows if these observations represent important human pre-

ferences in grammatical structure or if they are merely historical

accidents resulting from the survival and spread of some languages

at the expense of others.

See also: typology

Further reading: Comrie 1989; Whaley 1997.

BEHAVIOURISM

The view that psychology should invoke only observable and mea-

surable phenomena. Early in the twentieth century, psychology had

become somewhat obscurantist and even metaphysical. Behaviour-

ism originated as a healthy reaction to this state of affairs: the early

behaviourists wanted to sweep away what they saw as empty spec-

ulation and the endless postulation of undetectable concepts. They

BASIC WORD ORDER

30

therefore resolved to deal with nothing except what could be

directly observed and preferably measured. Along with their rejec-

tion of the excess baggage of earlier approaches, they often went so

far as to reject such intangible concepts as ‘emotions’, ‘intentions’,‘purposes’ and even ‘minds’.

Behaviourism exercised great influence over the linguist Leonard

Bloomfield and the American structuralists who followed him: they,

too, preferred to concentrate on directly observable linguistic

behaviour and to refrain from abstract theorizing.

In 1957 the American psychologist B.F. Skinner published Verbal

Behavior, an attempt at interpreting language acquisition strictly in

terms of behaviourism, and by far the most radical attempt ever attreating language in a behaviourist framework. Skinner’s book was

savagely (some would say unfairly) reviewed by the young Noam

Chomsky, who argued vigorously that Skinner’s approach not only

explained nothing but could not possibly explain anything of

interest. (More recent work on acquisition has reinforced Choms-

ky’s arguments by demonstrating that first-language acquisition is

clearly not, as Skinner had maintained, an essentially passive affair,

but that young children actively construct their language as theygo). Fair or not, Chomsky’s review persuaded a whole generation of

linguists that the essentially atheoretical behaviourist approach had

nothing to offer linguistics. As a result, the linguists influenced by

Chomsky abandoned behaviourism and embraced mentalism instead,

and linguistics was eventually integrated into the emerging dis-

cipline of cognitive science.

See also: language acquisition; language instinct; mentalism

Further reading: Malmkjær 2004; Pinker 1994; Trask 1995.

BILINGUALISM

The ability to speak two languages. In modern western society, the

ability to speak two languages is often seen as something of aremarkable achievement, particularly in the English-speaking

countries. However, over 70 per cent of the Earth’s population are

thought to be bilingual or multilingual (able to speak three or more

languages), and there is good reason to believe that bilingualism or

multilingualism has been the norm for most human beings at least

for the past few millennia. There is evidence that children raised

BILINGUALISM

31

bilingually tend to be more expressive, more original and better

communicators than children raised with only one language.

In New Guinea, in south-east Asia, in India, in the Caucasus, in

the Amazon rainforest, people routinely learn two or three neigh-bouring languages as well as their own, and the same was true of

Australia before the European settlement. Even today, many mil-

lions of Europeans are at least bilingual, speaking both their own

mother tongue and the national language of the country they live

in, and many of them can additionally speak a global language or

world language like English or French.

Bilingualism can be the property of an individual, but equally it

can be the property of an entire speech community in which two ormore languages are routinely used. The existence of bilingual and

multilingual societies raises a number of important social, political

and educational issues. In what languages should education be

delivered, and at what levels? What languages should be accepted

for publication and broadcasting? In what languages should laws be

written, and what languages should be accepted in court proceed-

ings? Differences of opinion may lead to language conflict, as

speakers jostle for the right to use their own mother tongues in thewidest possible domain, possibly to the exclusion of other languages.

It is also possible for an individual to speak two different dialects

of a single language – for example, her own regional dialect and the

standard language. In this case, we speak of bidialectalism.

See also: code-switching; diglossia; minority language

Further reading: Bonvillain 1993; Crystal 1997; Edwards 1994, 2004;Heller 2007; Li 2000; Romaine 1995; Steinberg 1993; Zentella 1997.

BIOPROGRAM HYPOTHESIS

The hypothesis that human beings are born with a ‘default’ struc-

ture for language built into their brains. Human languages differ

rather substantially in their grammatical structures (for example, intheir basic word order). However, creoles all over the world appear

to be strikingly similar in their grammar: all creoles look pretty

much alike, regardless of where they came into existence or of

which languages provided most of the input into them.

The linguist Derek Bickerton has proposed an explanation for

this observation. Since creoles are newly created languages, built

BIOPROGRAM HYPOTHESIS

32

out of the reduced pidgins which preceded them, and since the

children who create a creole are obliged to build its grammar for

themselves, Bickerton argues that there must be some kind of

innate machinery which determines the nature of that grammar. Hecalls this machinery the bioprogram, and he sees the bioprogram as

an innate default structure for language which is always imple-

mented by children unless they find themselves learning an adult

language with a different structure, in which case they learn that

instead. The bioprogram hypothesis therefore represents a rather

specific and distinctive version of the innateness hypothesis.

See also: genetic hypothesis of language; innateness hypothesis

Further reading: Bickerton 1981, 1984; Holm 1988–9, 2000.

BLACK ENGLISH

The distinctive varieties of English used by many native speakers of

African or Caribbean origin, and now more usually called African-American (Vernacular) English. Several important English-speaking

countries have sizable populations of black people whose ancestry

lies largely in Africa or the Caribbean; most prominent here are

Britain and the USA. For various historical reasons, the majority

of British and American black people, as well as many ethnically

white people living in connected areas of the southern and eastern

USA, speak varieties of English which are quite distinctive, differ-

ing from other varieties in vocabulary, pronunciation, grammar andmodes of discourse.

These varieties of English exhibit characteristics which are

often of considerable linguistic interest: for example, many have

highly unusual verbal systems which make distinctions unknown

in other types of English. It is possible that some of these fea-

tures may continue characteristics of the creoles once used by the

ancestors of the present-day speakers, and a few may even con-

tinue features present in the mother tongues of Africans sold intoslavery long ago.

But the primary reasons for the recent interest in black varieties

are not linguistic, but rather social, political and educational. Like

speakers of other distinctive varieties, speakers of Black English

often regard their mother tongue as a badge of identity and a

matter of pride: abandoning it may be seen as an act of betrayal. At

BLACK ENGLISH

33

the same time, just as with other groups, failure to acquire a com-

mand of standard English is a serious obstacle to making a career

in all but a few professions, and Black English itself may be

strongly stigmatized among white speakers in formal settings,though it has been extremely influential and prestigious amongst

younger speakers of English of all races in Britain and the USA.

Consequently, politicians, academics, teachers and school admin-

istrators, both black and white, are faced with some difficult ques-

tions of how to regard Black English. Some people advocate the

extreme position of recognizing and teaching only standard English

and of attempting to stamp out Black English. Most linguists, and

some others, would see this stance as unworkable and destructive,and would advocate the encouragement of bidialectalism: compe-

tence in both Black English and standard English.

Recently, however, a number of influential commentators, parti-

cularly in the USA, have been vigorously advocating another posi-

tion: Black English, now renamed Ebonics, should be recognized

not only as the equal of standard English but even as a totally

separate language, and it should be the language of instruction in

schools and even a school subject itself, a policy already adopted(though only briefly) by at least one American school board. These

debates continue, not only amongst politicians and observers but

within sociolinguistics.

See also: creole; sociolinguistics

Further reading: Holmes 1992; Llamas et al. 2007.

BORROWING

Whenever speech communities come into contact, languages often

take elements from each other, even if the extent of their mutual

influence does not develop further (into pidginization, for example).

Such borrowing is most commonly at the lexical level: words are

borrowed where the source language has lexicalized a concept thathas become useful in the target language. Examples from English

include the useful words schadenfreude (from German), chic (from

French), spaghetti (from Italian), juggernaut (from Persian), curry

(from Tamil), and many others. Sometimes these words retain their

foreign pronunciation and connotations; sometimes they are thor-

oughly assimilated. Innovations in politics, science, technology,

BORROWING

34

education, culture, emigration and so on are often signalled by

loan-words being borrowed from other languages: Latin and Greek

for scientific terms, French for culinary and administrative terms,

for example. Grammatical words (like modal auxiliaries, pronounsand determiners) are very rarely borrowed, and the borrowing of

syntactic structures is also unusual; where they are borrowed, it is

usually an indication of a major sociolinguistic upheaval in the

history of the language (as when English lost its Germanic case

system, or borrowed many words from French for which words

already existed in English).

See also: language change; language contact

Further reading: Crowley 1997; Fennell 2000.

CASE

The grammatical category by which the form of a noun phrase

varies for grammatical or semantic reasons. Case is indicated mor-phologically usually in the form of affixing to nouns; where such

case-marking exists, the language relies less on word-order to sig-

nify relations between the participants in the verbal process. Affixes

indicating case are called inflections. Not all languages have case,

but quite a few do. Consider Basque. The Basque noun phrase

etxea ‘the house’ has a number of different case-forms, such as:

A case-language must have at least two case-forms; most have

three to six distinct cases, and some, like Basque and Finnish, have

a dozen or more. English in its Anglo-Saxon and early Middle

English varieties used to distinguish nominative (subject), accusative

Name Form Function

Absolutive etxea intransitive subject; direct objectErgative etxeak transitive subjectDative etxeari ‘to the house’ (abstract relation)Genitive etxearen ‘of the house’ (possessor)Instrumental etxeaz ‘by means of the house’Comitative etxearekin ‘with the house’Locative etxean ‘in the house’ (location)Ablative etxetik ‘from/out of the house’Allative etxera ‘to the house’ (motion)

CASE

35

(direct object), genitive (possessive relations) and dative (abstract

relations) by attaching case-suffixes to nouns. Now English only

has remnants of its case ancestry, in the pronoun forms I/me, yours/

mine, he/him/his and she/her/hers, in who and whom, and in thegenitive case-ending -es which is now ellipted with an apostrophe:

Peter’s book, the reader’s eyes.

Many different cases have been identified in the world’s lan-

guages. Here are the most common and their typical functions,

though these vary across different languages:

Nominative: subject participant, these nouns are typically unmarked

or have a null-inflection.Accusative: the direct object.

Dative: indirect object.Genitive: indicating a possessive relationship between one noun and

another.

Ablative: indicating motion or movement, sometimes called the

adverbial case.

Vocative: used where the noun is a direct address.

Abessive: used to indicate absence, as in the Finnish inflection -tta:rahatta ‘without money’.

Comitative: shows accompaniment or instrumentality, as in the

Estonian -ga: ninaga ‘with a nose’.

Allative: indicates motion or positioning ‘onto’ or ‘on top of’.

Essive: used for a temporary state, as in the Finnish -na: pikku-

lapsena ‘as a baby’.

Temporal: used to indicate points in time, as in the Hungarian -kor:

karacsonykor ‘at Christmas’.Aversive: where the noun is something to be feared.

A special case is the ergative, which defines the relationship of

agency between nouns in a sentence. A very general distinction can

be made between ergative languages and nominative languages,

based on how they treat these differences in transitivity. Roughly,

ergative languages focus their articulation on the agency of the

utterance, while nominative languages focus on the subject of thesentence. More specifically, in nominative languages like English the

subject of a sentence with a transitive verb and the subject of a

sentence with an intransitive verb are treated alike, while direct

objects of transitive verbs are treated differently; in ergative lan-

guages, the subject of an intransitive verb is treated the same as the

direct object of a transitive verb, and it is the subject of transitive

CASE

36

verbs that are treated differently. In English for example, the

grammar in the two sentences Helen opened the door and The door

opened is quite different, though the agency of the event might be

thought of as being the same. A language with an ergative casewould articulate these relationships very differently. Examples of

ergative languages include Basque, Inuit, Kurdish, Tagalog, Tibetan

and many native Australian languages like Dyirbal.

In the Government-and-Binding Theory, the idea of case has been

generalized and made abstract, and a Case (with a capital letter) is

assumed to belong to every noun phrase in every grammatical sen-

tence. In this approach, the linguistic realization of Case in the

surface structure of the language is a matter of the transformationalrules local to the language.

See also: morphology; transitivity

Further reading: Blake 2001; Dixon 1994; Hurford 1994.

CLAUSE

The largest grammatical unit smaller than a sentence. The clause is

a traditional and fundamental unit of sentence structure, though

the term is not used by all grammarians in exactly the same way.

Traditionally, a clause is a grammatical unit consisting of a subject

and a predicate, and every sentence must consist of one or more clau-

ses. In the following examples, each clause is marked off by brackets.

A simple sentence consists only of a single clause: [Susie has bought

a skirt].

A compound sentence consists of two or more clauses of equal rank,

usually joined by a connecting word like and, or or but: [Susie wants

children], but [her career won’t allow them].

A complex sentence consists of two or more clauses where one out-

ranks the others, which are subordinated to it: [After she got her

promotion], [Susie bought a new house].

A clause which is the highest-ranking, or only, clause in its sentence

is a main clause; a clause which is subordinated to another is a

subordinate clause. Traditional grammarians usually regarded a

subordinate clause as entirely separate from the higher-ranking

clause it is attached to, but today linguists normally regard a

CLAUSE

37

subordinate clause as forming a part of its higher clause, and the

examples below show this.

There are several types of subordinate clause:

An adverbial clause is related to its higher clause like an adverb:

[Susie develops a rash [whenever she eats strawberries]].

A complement clause is attached to a preceding word (usually a verb

or a noun) and ‘completes’ the sense of that word: [Susie has deci-

ded [that she will look for a new job]]; [The rumour [that Susie is

quitting] is not true].

An embedded question is a question buried in a larger sentence:

[Susie has not decided [what she is going to do]].A relative clause modifies a noun: [The skirt [that Susie bought] is

too short].

Recently some grammarians have been extending the term clause toevery unit containing a verb, including many units traditionally

regarded only as phrases. Examples: [Susie’s heavy smoking] is

affecting her health; Susie wants [to buy a new car]; [Having finished

her dinner], Susie reached for her cigarettes. This extended usage isnot standard, but it is now very widespread. Note that the brack-

eting system adopted here is convenient for linear text, but expres-

sing syntax in a tree diagram is much clearer and more common.

See also: phrase; sentence; subordination; tree

Further reading: Collins Cobuild 1990; Crystal 1996; Hurford 1994;Jackson 2002; Jeffries 2006.

CLITIC

A clitic is attached to a word (its host) and the two are pronounced

together but the two are treated separately at the level of syntax.

For example, the contractions ’s, ’ll and ’m in he’s running, I’ll be

there, and I’m not playing are all clitics. A clitic is never stressed,even when its host is, which is why it has been argued that n’t

(couldn’t, won’t) is not a clitic, since the not of the full form (could

not, will not) can never be unstressed. Determiners (the pronounced

both as /D e/ and /Di/, a and an) are also treated as clitics, since they

cannot appear on their own but must accompany a head noun.

Those which precede the host are known as proclitics; those which

CLITIC

38

follow the host are enclitics; and those (as in Portuguese) which are

inserted into the word are mesoclitics.Cliticization in different languages provides important data on

the rules of constituent structure and phonology.

See also: auxiliary; syntax

Further reading: Heggie and Ordonez 2004.

CLOZE

A cloze test or cloze procedure was a popular method in languagetesting, which involved presenting learners with a text from which

certain words had been deleted. The student must then fill in the

blanks in a way that preserves linguistic well-formedness anddemonstrates comprehension. It was popular in the 1970s and

1980s since it was easy to design and assess (it could even be auto-

mated), and it could be graded for difficulty from the deletion of

simple content words to test understanding, or the deletion of

words requiring grammatical agreement or declension, up to the

deletion of auxiliaries, determiners or prepositions to test precise

linguistic competence in the second language. However, it came to

be criticized on the basis that it actually tested clause-level abilityrather than text-level skills, and the cloze-spaces could be guessed

at more readily by students with cultural framing knowledge rather

than linguistic knowledge. It was also often the case that native-

speakers did not score 100 per cent, since they were using more

idiomatic or register choices that evaded the cloze design.

Cloze tests continue to be used in the second-language class-

room, nevertheless. They have also proven useful in psycholinguis-

tics as tests of comprehension, and in sociolinguistics as a means ofeliciting dialect words from informants.

See also: applied linguistics; sociolinguistics

Further reading: Buckingham and Yorkey 1984; Oller and Jonz 1994.

CODE-SWITCHING

Changing back and forth between two language varieties, especially

in a single conversation. Sociolinguists use the term code to denote

CODE-SWITCHING

39

any identifiable speech variety, including both a particular language

and a particular variety of a language. Many speakers have control

over at least two varieties of their language (for example, a

German-speaker may speak both his local variety of German andstandard German), and many more have control over two lan-

guages (for example, Welsh/English bilinguals in Wales). Such

speakers will shift back and forth between these varieties, depend-

ing on such factors as who they are talking to, where they are, and

what they are talking about. This is code-switching.Speakers switch between languages, dialects, styles or accents

during a conversation usually because the setting itself has changed

and demands a different variety: this is situational code-switching.Alternatively, speakers can deliberately switch to another code in

order to force a new perception of the situation on the other parti-

cipants: this is called metaphorical code-switching. Switching can

happen at thematic boundaries in a conversation, between sentences,

between clauses, or even as tag-phrases at the end of utterances.

See also: bilingualism

Further reading: Auer 1998; Crystal 1997; Holmes 1992; Li 2000;Romaine 1994; Stockwell 2002a.

COGNITIVE LINGUISTICS

An approach to the study of language which is based upon human

perception and conceptualization of the world. During the twen-tieth century the most influential approach to the study of language

has been structuralism: linguists have largely devoted themselves to

the purely structural aspects of language systems themselves, such

as sound systems and grammatical systems. A key feature of struc-

turalism is that it concentrates on the internal structure of a lan-

guage and not on the way in which the language relates to the non-

linguistic world.

Naturally, links between languages and the world have not beenneglected, and anthropological linguistics in particular has been

devoted to studying links between language and culture. Since

about 1980, however, a growing number of linguists have been

devoting serious attention to a more ambitious project: the eluci-

dation of the ways in which linguistic objects and structures reflect

the manner in which human beings perceive, categorize and

COGNITIVE LINGUISTICS

40

conceptualize the world. This is cognitive linguistics: together with

cognitive psychology and neurolinguistics they constitute the field of

cognitive science.Among the early contributors to the cognitive approach was the

American theoretical linguist George Lakoff, who has written

extensively on the importance of metaphor in shaping languages.

More recently, a number of people with diverse backgrounds have

been attempting to analyse linguistic structures in terms of concep-

tual and perceptual categories like figure and ground, landmark and

trajector (something which moves), location in space, events and

states, frames and schemas (scripts or mental models of real and hypo-

thetical world objects and events), and categories and hierarchies.The American linguist Ronald Langacker has constructed a theory

of grammar, called cognitive grammar, on the basis of these ideas.

See also: foregrounding; metaphor; mentalism; schema

Further reading: Hickmann and Robert 2006; Hiraga 2005; Hofstadter1979; Johnson-Laird 1983, 1993; Kovecses 2005; Lakoff 1987; Lakoffand Johnson 1980; Langacker 1987–91, 1990; Palmer 1996; Stillings etal. 1987; Stockwell 2002b; Ungerer and Schmid 1996.

COHESION

The presence in a discourse of explicit linguistic links which provide

structure. Among these devices are anaphors like she, they, this and

one another, temporal connectives like after and while, and logicalconnectives like but and therefore. Every one of these items serves

to provide some kind of specific link between two other smaller or

larger pieces of discourse.

Consider a pair of examples. In the first, the cohesion has gone

wrong: The Egyptians and the Assyrians were carrying standards

some 5,000 years ago. They were poles topped with metal figures of

animals or gods. Here the reader might naturally take they as refer-

ring to the Egyptians and the Assyrians, and is likely to be flum-moxed by the continuation. The second version is different: Some

5,000 years ago, the Egyptians and the Assyrians were carrying

standards. These were poles topped with metal figures of animals or

gods. This time the item these immediately makes it clear that it is

the standards which are being referred to, and the continuation is

smooth and effortless.

COHESION

41

Naturally, the proper use of cohesive devices has long been

recognized as a fundamental aspect of good writing, but in recent

years linguists have been turning their attention to the analysis of

these devices. The term cohesion was coined by the British linguistMichael Halliday, and the study of cohesion is especially prominent

within Halliday’s Systemic Linguistics, but it is also now a familiar

part of most linguistic analyses of texts and discourses. An impor-

tant aspect of this is lexical cohesion, which describes the ways in

which words are related semantically or in relation to the experi-

ential schemas which people accumulate through their lives.

Note that cohesion and coherence are not necessarily the same

thing. A text can be lexicogrammatically cohesive while beingutterly incoherent for understanding. Cohesion is a linguistic

property, while coherence is a combined linguistic and cognitive

property.

See also: Systemic Linguistics; text

Further reading: Halliday 2004; Halliday and Hasan 1976; Mastermanand Wilks 2005, Thompson 2004; Wolf and Gibson 2006.

COLLIGATE

Two words display colligation if they are related syntagmatically

and grammatically: in other words, if they occur close to one

another and are significant at the grammatical level. For example,

infinitive forms of a verb would colligate with each other, presentprogressive -ing forms would colligate together, phrasal verbs which

take the particle up (run up, throw up, catch up) would all be colli-

gates. The term is usefully contrasted with collocation which is a

lexical rather than grammatical relationship. For example, fat typi-

cally collocates with man, lady, chance, cheque but not with brick,

wine or car. This is not to say that the latter sort of collocations

never occur, but collocation is a relatively predictable relationship,

and the range of collocational possibilities is known as a word’scollocability. Both of the terms colligate and collocate are used

extensively in corpus linguistics.

See also: computational linguistics; semantics

Further reading: McEnery and Wilson 2001; McEnery et al. 2006.

COLLIGATE

42

COMMUNICATIVE COMPETENCE

The ability to use language appropriately in social situations. In

order to speak a language successfully, you need to have purelylinguistic competence in that language: mastery of pronunciation, of

grammar and of vocabulary. But you need more than that: you also

need sociolinguistic competence, knowledge of such things as how to

begin and end conversations, how and when to be polite, and how

to address people. In addition, you further need strategic compe-tence, knowledge of how to organize a piece of speech in an

effective manner and how to spot and compensate for any mis-

understandings or other difficulties.The concept and the term were introduced by the American lin-

guist Dell Hymes in the 1970s. Hymes was dismayed by what he

saw as the excessively narrow concern of many linguists with noth-

ing but internal linguistic structure, at the expense of communica-

tion, and he wished to draw attention to the importance of

appropriateness in language use.

Today linguists of a theoretical orientation still prefer to focus on

the purely structural aspects of language, but those with an interestin anthropological linguistics, in functionalism, in sociolinguistics, in

language teaching, or in communication generally typically attach

great importance to the examination and elucidation of commu-

nicative competence.

See also: ethnography of communication; sociolinguistics

Further reading: Bonvillain 1993; Foley 1997.

COMMUNITY OF PRACTICE

A notion from sociolinguistics which goes beyond the definition of

a speech community by encompassing all the cultural and

behavioural practices of a group (including linguistic practices),

which together serve to define that group. The idea has been par-ticularly fruitful in relation to the study of gender and language,

since it takes a contextual view of the power relations and

ideologies which influence the language practices within a group.

The CoP approach also suggests that linguistic behaviour is adap-

table and variable when the individual moves between different

groups in her life, and so it has refocused linguists’ minds

COMMUNITY OF PRACTICE

43

on the online, provisional and transitory nature of much language

usage.

See also: gender; social network; sociolinguistics; speech community

Further reading: Eckert and McConnell-Ginet 1992, 1995, 1999, 2003;Wenger 1998.

COMPARATIVE RECONSTRUCTION

The principal method used to find information about an unattested

language which is the ancestor of several known languages. Every

living language changes over time, but it does not change everywhere

in the same way. A language spoken over a sizable area may thereforebreak up, first into regional dialects, and eventually into several quite

diverse daughter languages. When we find several languages, either

spoken today or abundantly recorded in written texts, which clearly

share a common ancestor, we have ways of working backwards to

figure out what that unrecorded ancestor was like. Chief among

these is comparative reconstruction, or the comparative method.The comparative method deals with the phonological forms of

words, and its successful use in a given case depends upon the cor-rectness of three assumptions. First, we assume that a significant

proportion of the words of the ancestral language still survive in

the recorded daughter languages. Second, we assume that those

surviving words in most cases have not changed their meanings too

dramatically. Third, we assume that phonological change (change in

the pronunciations of those words) is generally regular – that is,

that a given sound in a given environment has consistently changed

in exactly the same way in a single daughter language. Only whenthese three assumptions are substantially correct can we apply the

comparative method.

Below is a modest example, showing the words for certain

meanings in four European languages for which we have excellent

reasons for believing that they share a common ancestor.

English Latin Greek Irish Gaelic

fish piscis ikhthys iasgfather pater pater athairfoot ped- pod- troigh


44

The key point here is the systematic correspondences which we

can see in many cases. For the first set, we observe that a native

English word beginning with /f/ is matched by a Latin word begin-

ning with /p/ and a Greek word beginning with /p/. The Irish case is

a little more difficult, but examination of further data reveals thatthe usual Irish correspondence here is zero: no initial consonant at

all, as in the words for ‘fish’ and ‘father’. The reason that the Irish

words for ‘foot’ and ‘for’, and also the Greek word for ‘fish’, do not

match is that the ancestral words for these senses have been

replaced in these cases by different words. (Remember our first

assumption: not too many ancestral words have been lost and

replaced.)

Likewise, in the second set, we find initial /s/ consistently in allthe languages except Greek, where instead we find initial /h/ with

equal consistency. Again, the word for ‘sweet’ has been lost and

replaced in Irish.

The third set exhibits initial /n/ in all four languages, though the

Greek word for ‘nine’ has acquired a prefix absent elsewhere. The

Irish word for ‘night’ does not fit, but the Irish word for ‘tonight’

does fit, and we may reasonably assume that the word for ‘tonight’

contains an earlier word for ‘night’, now itself lost and replaced.(Recall our second assumption: we can tolerate a certain amount of

change in meaning, but not too much.)

Now the explanation for these correspondences resides crucially

in our third assumption: change in pronunciation has mostly been

regular in all four languages. So, we conclude, the words in the first

group all began with the same sound in the ancestral language, and

that sound has developed regularly into /f/ in English, into /p/ in

Latin and Greek, and into zero in Irish. Similarly, all the words inthe second group began with a different same initial sound in the

ancestral language, one which has regularly developed into /h/ in

for pro para dosix sex hexa seseven septem hepta seachtsweet suavis hedys milissalt sal hal salannnew novus neos nuanight noct- nykt- (in)nocht ‘tonight’nine novem (en)nea naoi

English Latin Greek Irish Gaelic


45

Greek but into /s/ in all the others. And all the words in the third

group began with yet another same sound in the ancestral language,

one which has developed into /n/ in all four languages.

It remains only to decide what the ancestral sounds were in eachcase, and this is the central step in comparative reconstruction. By

examining all the available evidence, and by knowing what kinds of

phonological changes are frequent and natural in languages, spe-

cialists have determined that, in the ancestral language, all the

words in the first group originally began with */p/, those in the

second with */s/, and those in the third with */n/. (The asterisk

denotes a reconstructed sound.) The original */p/ has changed to /f/

in English and been lost in Irish, and the original */s/ has changedto /h/ in Greek, all of these being changes which are frequent in

languages and easy to understand. Apart from these cases, the three

ancestral sounds, in initial position, have remained unchanged in all

four languages.

Of course, we do not apply the comparative method only to

word-initial sounds: we must figure out, as far as possible, what the

entire words looked like in the ancestral language. Specialists have

managed to determine that the forms of these words in the ances-tral language were approximately as follows (endings omitted):

*pisk-, * p

e

ter, *ped-, *per, *sweks, *septm, *swad-, *sal-, *newo-,

*nekwt- and *newn. The change in pronunciation in each of the

four daughter languages has been largely regular, though naturally

with a few complications here and there. And the unrecorded

ancestral language in this case, of course, is Proto-Indo-European,the ancestor of the Indo-European family, to which all four lan-

guages belong.The comparative method consists of the entire business of

deciding that certain languages probably share a common ancestor,

identifying systematic correspondences, and working backwards to

identify the forms of the words in the ancestral language, and this

method is the cornerstone of work in historical linguistics. If we

can’t find systematic correspondences, then we can’t do compara-

tive reconstruction, and any miscellaneous resemblances that we

come across are very likely only the result of chance, or perhapseven of ancient language contact.

See also: internal reconstruction; reconstruction; systematic corres-pondence

Further reading: McMahon 1994, 2000; Trask 1996.


46

COMPETENCE

Our ability to use language, viewed in the abstract. When we use

our language, we commit all sorts of errors. We make slips of thetongue, we sometimes can’t think of a word or name we know per-

fectly well, we interrupt ourselves, we mishear or misunderstand

what others are saying, we may even lose the thread of what we

ourselves are saying, there are limits upon our memories, and so

on. There are also sociological factors that affect the language we

use. In the early 1960s, the American linguist Noam Chomsky

began arguing that such errors should be dismissed from con-

sideration in examining our ability to use language. Chomskyargued that every one of us possesses an abstract linguistic compe-tence which is independent of the errors we sometimes make, and

he argued further that the elucidation of this competence was, or

should be, the principal business of linguistic theory. The errors and

social context he relegated to the quite different domain of perfor-

mance, which he considered to be the proper subject-matter of

different disciplines.

Chomsky’s position here has been enormously influential in lin-guistics, and it has formed the basis of his research programme,

dedicated to the identification of the highly abstract principles

which he sees as making up our competence. Interestingly, Choms-

ky’s distinction is strikingly reminiscent of the distinction between

langue and parole introduced generations earlier by the Swiss lin-

guist Ferdinand de Saussure, though not quite identical to it.

Nevertheless, there are critics who see Chomsky’s conception of

competence as far too abstract, and who are inclined to doubtwhether such a sharp line should or can be drawn between our

ability to use language and our actual behaviour. Sociolinguists,

applied linguists and cognitive linguists, for example, are interested

in what would be classified here as aspects of performance.

See also: langue; language faculty; performance; universal grammar

Further reading: Harris 1993; Newmeyer 1983; Matthews 1979.

COMPLEMENT

A grammatical unit which contains a verb and which forms part of

a larger unit. The term complement has in fact been used by

COMPLEMENT

47

linguists in a variety of senses, but it is most commonly applied to a

grammatical unit which contains a verb and which, in some sense,

completes a larger grammatical unit which begins with some other

words.Consider the sentence The rumour that John is a Chinese spy

amuses me. Here everything before amuses is a single noun phrase,

the subject of the sentence, and it contains a noun-complementclause, which is that John is a Chinese spy; this is attached to the

noun rumour, and it completes the subject noun phrase.

Slightly different are Susie told me that she would come and I

don’t know whether she’s coming, in which that she would come and

whether she’s coming are verb-complement clauses attached to theverbs told and know, respectively. The items that and whether, which

introduce complement clauses, are assigned to a part of speech

called complementizers. (Traditional grammarians called them

conjunctions.)

The complements illustrated above are all finite. But we also have

non-finite complements: the sentence Susie wants to buy a car con-

tains the non-finite complement to buy a car, which is attached to

the verb wants. Note, however, the big difference between Susie

wants to earn some extra money, in which to earn some extra money

is a direct complement attached to wants, and Susie moonlights to

earn some extra money, in which to earn some extra money is a

purpose complement; the second, but not the first, is equivalent to in

order to earn some extra money.

The term complement is also applied to a variety of other things

which appear to ‘complete’ a sentence in some sense. For example,

in Susie is considered clever, the item clever is sometimes called asubject-complement (it applies to Susie, the subject of the sentence),

and, in Susie finds Mike tiresome, the item tiresome is sometimes

called an object-complement (it applies to Mike, the object of the

verb).

See also: argument

Further reading: Greenbaum and Quirk 1990; Hurford 1994.

COMPUTATIONAL LINGUISTICS

The use of computers to perform various tasks involving language.

The introduction of digital computers has made possible a number

COMPUTATIONAL LINGUISTICS

48

of approaches to descriptive and practical problems of language

which could not previously be addressed adequately or at all.

One obvious use of computers is to store a corpus of spoken or

written texts. Such a machine-readable corpus can be rapidly sear-ched and interrogated in order to obtain such information as the

frequency of occurrence of certain words, forms or constructions. In

this way we can obtain hard data about real language use which

would not otherwise be accessible, and we can further make com-

parisons between, say, spoken and written English, or between

American and British English. This is corpus linguistics, which has

had enormous success in recent years in the creation of new gram-

mars based on spoken usage as well as on writing. Concordances areused extensively in the field: a concordance is a computed list of

every occurrence of an individual word in the corpus. A corpus can

be anything from a small set of texts parsed and tagged right up to

multi-million word digitized collections of spoken and written data.

Both grammarians and lexicographers today routinely base their

grammatical descriptions and their dictionaries on the data extrac-

ted from vast corpora.

Computational linguistics has also been influential in machinetranslation (the development of computer programs which can take

a text written in one language and convert it into a different lan-

guage), speech synthesis and voice recognition (converting written input

into an intelligible imitation of human speech), computer-assistedlanguage learning (CALL), and in the stylistic analysis of texts.

See also: colligate; natural-language processing

Further reading: Barnbrook 1996; Biber 2006; Crystal 1997; Garside etal. 1997; Mahlberg 2005; McEnery and Wilson 2001; McEnery et al.2006; O’Grady et al. 1996; Sampson and McCarthy 2005; Sinclair1991; Sinclair and Carter 2004; Tognini-Bonelli 2001.

CONCEPTUAL INTEGRATION

A theory within cognitive linguistics which places the human

capacity for dealing with metaphor at the heart of our linguistic

ability. Developed in the late 1990s by Gilles Fauconnier and

Mark Turner out of Fauconnier’s earlier work in mental spaces, CItheory explains individuals’ ability to understand novel situations

which go beyond anything comparable in their experience. Early

CONCEPTUAL INTEGRATION

49

cognitive linguistic theory saw a metaphor as a mapping from a

source domain of familiar semantic content onto a target domain of

unfamiliar and new information (JULIET IS THE SUN, LIFE IS A JOURNEY).

This target domain was then restructured in terms of the source, toproduce a new (metaphorical) understanding. However, it became

clear that the products of such mappings often went beyond a

simple restructuring on the basis of existing knowledge: domains

were blended to take on an emergent logic of their own.

The theory has been productive especially in relation to creative,

literary and expressive metaphors in the field of cognitive poetics.

See also: cognitive linguistics; metaphor

Further reading: Fauconnier and Turner 2002; Stockwell 2002b.

CONJUNCTION

The part of speech which includes words like and and or. Today the

label conjunction is normally only applied to a very small group ofwords, chiefly and and or, which were traditionally called the coordi-nating conjunctions. Most usually, a conjunction conjoins (joins) two or

more instances of the same category. Examples: Would you like a

doughnut or a piece of pie? (conjoined noun phrases); Susie sipped her

drink, lit a cigarette and opened her book (conjoined verb phrases).

Traditional grammarians also included among the conjunctions

another group of words, the subordinating conjunctions or sub-ordinators. These are the words like if, whenever, although and after,which introduce adverbial clauses. Examples: After she finished her

essay, she headed for a shower; If she arrives in time, she can come with

us. But these words behave very differently from the true conjunc-

tions, and today they are normally placed in a class by themselves.

Traditional grammarians also counted as conjunctions the com-plementizers like that and whether, as in She said that she would

come and I don’t know whether she’s coming. But these words

behave differently again, and today they are placed by themselves inyet a third class.

See also: coordinate structure

Further reading: Collins Cobuild 1990; Crystal 1996; Hurford 1994;Jeffries 2006.

CONJUNCTION

50

CONNOTATION

The meaning of a word that is broader than its central and primary

sense, often acquired through frequent associations. The wordrugby has as its central sense (its denotation) a particular type of

football game, but, depending on your experience of rugby, it may also

conjure up in your mind such associations as ‘large men’, ‘manliness’,

‘boorish and bawdy behaviour’, or ‘public schools’ (that is, expen-

sive and prestigious private schools); it may remind you of your

pride in your local or national team, or it may remind you of a

present or former boyfriend. All these associations are part of the

connotation of the word. Even more loose are the meanings byassociation, which tend to be even more personal and idiosyncratic.

See also: denotation; meaning

Further reading: Jeffries 2006.

CONSONANT

A speech sound produced by significantly obstructing the flow of air

through the vocal tract. All speech sounds are produced by manip-

ulating the vocal tract while air is flowing through it. If this

manipulation does not produce any significant obstruction of the

airstream, the result is a vowel; if it does produce obstruction, the

result is a consonant.Consonants are classified into several types, differing in the kind

of obstruction involved. If the vocal tract is blocked completely,

and then the closure is released suddenly, the result is a plosive, like[p] or [d]. If the vocal tract is blocked completely and the closure is

released slowly, producing friction noise, the result is an affricate,

like [ts] or [S]. If the vocal tract is not completely blocked, but is

reduced instead at some point to a tiny opening through which the

air is forced, producing friction noise, the result is a fricative, like [f]or [z]. These three types together are obstruents; all other con-sonants are resonants.

If a complete closure is made in the mouth, but the velum is

lowered so that air can flow out freely through the nose, the result

is a nasal, like [m] or [n]. If a complete closure is made along the

centre-line of the vocal tract, but space is left for the air to flow out

along one or both sides, the result is a lateral, like [1]. If a speech

CONSONANT

51

organ is ‘flicked’ rapidly against another one, touching it briefly and

then returning immediately to its starting point, the result is a tap,like the sound [Q] in Spanish pero ‘but’. If a similar flicking move-

ment is made, but the moving organ ends up in a different placefrom where it started, the result is a flap, as in the unusual kind of

/r/-sound found in some languages of India. (Many books, espe-

cially in the USA, do not distinguish between taps and flaps, but it

seems helpful to do so.) If a small opening is made in such a way

that air forced through the opening causes some organ to vibrate

vigorously, the result is a trill, as in the [r]-sound of Spanish perro

‘dog’.

There is one more class of consonants, but these do not really fitour definition. The [j]-sound at the beginning of English yes, the

[w]-sound in wet, and most of the several different sounds used by

English-speakers to pronounce the /r/ of red are all strictly vowels

by the definition, since they involve no significant obstruction. But

the point is that these sounds behave like consonants in English and

other languages, in spite of their vowel-like nature, and hence they

are commonly regarded as a distinct group of consonants, the

approximants. See the remarks under vowel for an explanation.

See also: speech sound; vowel

Further reading: Ashby 1995; Ball and Rahilly 1999; Ladefoged 1993.

CONSTITUENT STRUCTURE

A type of hierarchical grammatical structure in a sentence. Con-

sider the sentence: The little girl washed her doll. In the view of most

linguists, this consists of two pieces, or grammatical units: the little

girl and washed her doll. The first of these in turn consists of the

plus little girl, and this last consists of little plus girl. The second

likewise consists of washed plus her doll; of these, the first consists

of wash plus -ed and the second of her plus doll.

This is the sort of grammatical (syntactic) structure exhibited byall sentences in English and in most other languages, and we call it

constituent structure. Constituent structure is hierarchical: a sen-

tence consists of a couple of large pieces, each of which consists of

some smaller pieces, each of which in turn consists of some still

smaller pieces, and so on, until we reach the smallest pieces of all, the

words or morphemes. And every one of these pieces is a constituent

CONSTITUENT STRUCTURE

52

of the sentence. Moreover, every constituent must belong to some

particular syntactic category: that is, the grammar of English (or of

any language) allows only constituents belonging to certain cate-

gories to be combined into certain larger categories. An attempt atusing a constituent of the wrong category produces an ungramma-

tical result, as in *The under the bed washed her doll (the asterisk

marks ungrammaticality).

Sentence structure, or syntax, is thus a very orderly affair: every

grammatical sentence is built up from smaller constituents com-

bined into larger constituents, according to certain rigid rules, the

rules of grammar for that particular language. The resulting struc-

ture of a sentence can be revealingly exhibited by a tree.Some languages, including many Australian languages, seem not

to have this sort of sentence structure; instead, they exhibit a much

looser type of structure in which smaller units do not have to be

combined into larger ones in such an orderly way. These are called

non-configurational languages, or W-star languages. Languages like

English, in contrast, are configurational languages.

See also: phrase; phrase-structure grammar; syntax; syntactic category;tree

Further reading: Brown and Miller 1991; Burton-Roberts 1986; Kroeger2004; van Valin 2001.

CONSTRUCTION GRAMMAR

A set of related approaches to grammar that treat constructions at

clause level as the units of grammar, rather than the syntactic unit.

The construction has an abstract conceptual content (such as ‘Xmoves over Y’ or ‘A causes B to die’), which is then related to var-

ious stylistic realizations: so construction grammars have been seen

as the point where generative semantics shifted into cognitive lin-

guistics in the 1980s. An example of a construction grammar would

be Ronald Langacker’s cognitive grammar which understands pre-

positions as being the surface realization of a set of spatial image-

schemas with a mental origin in physical experience.

Crucially, construction grammars are not derivational. So forexample, the active and the passive forms of a sentence are regarded

as having different conceptual structures rather than one being a

transformation of the other. Since construction grammars depend

CONSTRUCTION GRAMMAR

53

on the conceptual meaning in context, they can be seen as approa-

ches to linguistics that collapse the classical distinctions between

semantics, syntax and pragmatics. The construction is the unit of

language, which cuts across these other aspects. So, for example, inThey laughed him out of the room, the normally intransitive verb

receives a transitive reading and the situation can be interpreted on

the basis of the ‘X cause Y to move’ construction rather than the

syntactic deviance alone. As a result, construction grammars are

proving most useful in understanding language acquisition and are

being used for second-language teaching, since it is the mean-

ingfulness of the situation which is of primary importance, and

syntax and semantics are treated holistically.

See also: cognitive linguistics; language acquisition

Further reading: Croft and Cruse 2004; Goldberg 2006; Lakoff 1987;Langacker 1987–91.

CONTROL

The phenomenon in which a verb phrase (VP) with no subject is

interpreted as having some subject. In sentences like John wants to

go home and John promised Mary to go home (the second is not

grammatical for all speakers), we understand that it is John who is

going home. That is, John is the ‘understood’ subject of the sub-

jectless VP go home, and so we say that John, the subject of wants

and of promised, controls the VP go home. This is an example ofsubject-control. However, in the sentence John persuaded Mary to go

home, it is Mary who is going home; this time Mary, the object of

persuaded, controls the VP go home, and we have an instance of

object-control.There is one more possibility, illustrated by the sentence Going

home sounds like a good idea. This time there is no noun phrase

available at all to be interpreted as the subject of the VP going

home, and we call one an instance of arbitrary control. Controlphenomena can be rather intricate, and many current theories of

grammar devote a good deal of machinery to their treatment. Note

that some writers refer to control as sharing.Note that the term control has a second sense. In a number of

languages, intransitive sentences are divided into two different

grammatical types, distinguished by the degree of control exercised

CONTROL

54

over the action by the subject. In such languages, sentences like He

went home and He stood up, in which the subject has a high degree

of control, are expressed with one construction, while others, like

He died and He sneezed, in which the subject has a low degree ofcontrol, are expressed with a different construction.

See also: anaphor; raising

Further reading: Borsley 1991; Palmer 1994.

CONVERSATION ANALYSIS

A particular and highly empirical approach to examining the

structure of discourse. The term conversation analysis, which came

into prominence in the 1970s, sounds straightforward enough, but

it is used in two rather different ways. Some people use it in a very

broad sense, to include all possible approaches to the study of

conversational structure. Much more commonly, however, the term

is used more narrowly to denote an approach which rejects the useof traditional and widely used grammatical concepts and terms and

attempts instead to work out from observation what speakers are

doing and how they are doing it, with any required concepts and

terms being derived purely from observation. The leading figure in

the development of this approach was the American sociologist

Harvey Sacks.

The approach is particularly associated with a general approach

to social sciences called ethnomethodology, whose proponents arguethat the proper object of sociological study is the set of techniques

that the members of a society use to interpret their world and to act

within it. In practice, this means a minimum of theorizing and a

strong emphasis upon raw data and on the patterns that emerge

from the data. Consequently, conversation analysis, in this narrow

sense, contrasts most obviously with discourse analysis, which

operates from the beginning with the familiar concepts and terms

of general linguistics and attempts to examine the role of theseconcepts in discourses, including conversations.

See also: discourse analysis; turn-taking

Further reading: Duranti 1997; Eggins and Slade 2004; Levinson 1983;Mey 1993; Pridham 2001; Schiffrin 1994; Wooffitt 2005; Yule 1996.

CONVERSATION ANALYSIS

55

CONVERSATIONAL IMPLICATURE

A certain type of inference. Suppose Alice asks Bill Is Susie coming

to Mike’s party on Saturday?, and Bill replies Dave wants to go to a

concert. On the face of it, this is an idiotic response to a simple

question: Bill has declined to mention Susie at all and has instead

brought up Dave and a concert, neither of which was being asked

about. And yet this is a perfectly normal and satisfactory answer to

the question: providing that Alice knows that Dave is Susie’s boy-

friend, she can reason as follows: ‘Bill doesn’t know whether Susie

is coming to the party, or he would simply have told me, but Dave

is Susie’s boyfriend, and Bill tells me he wants to go to a concert;doubtless he will want Susie to come with him, and the concert

must be on Saturday, or Bill wouldn’t have mentioned it, and

therefore I can conclude that Susie will probably be going to the

concert with Dave, and hence that she won’t be coming to the

party’.

Alice’s conclusion that Susie probably won’t be coming to the

party is an example of a conversational implicature. This conclusionhas not been asserted by Bill, and it does not logically follow fromwhat Bill has said, and yet it is reasonable, and Alice will surely

draw it. How can this be so?

The first key point here is the context of Bill’s utterance. Alice

knows that Dave and Susie are a couple, and she knows that people

like their partners to accompany them to social events, or at least

that Dave does, and this contextual knowledge is crucial: without it,

Alice would have little chance of making sense of Bill’s response.

This is typically the case with an implicature: it can only be drawnby a hearer who has an adequate knowledge of the context.

A second key point is that Alice assumes that Bill is being coop-erative. If Bill had known for certain that Susie was or was not

coming to the party, Alice would have expected him to say so, and

failure to do this would be uncooperative. Moreover, Alice has

every right to assume that the concert in question must be on the

Saturday; had it been on the Friday, Bill’s behaviour would have

been very uncooperative indeed, not merely irrelevant but positivelymisleading. Alice therefore assumes that Bill is cooperating, and

draws her conclusion accordingly.

A notable property of implicatures is that they are defeasible –

they can be explicitly denied without producing anomaly. Suppose

Bill’s reply had been Well, Dave wants to go to a concert, but Susie

has decided to come to Mike’s party anyway. Here Bill is expressly

CONVERSATIONAL IMPLICATURE

56

denying the implicature ‘Susie probably won’t be coming to the

party’, but the result is still fine. This demonstrates that impli-

catures are not logically valid. Nevertheless, they are powerful

inferences, and an implicature which is not denied will be assumedby the hearer to be true.

Now it is certainly not possible to assert that Bill’s response

actually means ‘Susie is probably not coming to the party’. Suppose

Alice asks Bill a quite different question: I’d like to go to one of the

outdoor concerts in the park next week. Is anybody else interested?

Bill replies Dave wants to go to a concert. This time Alice is hardly

likely to conclude that Susie is probably not coming to Mike’s

party, since such an inference would make no sense in this verydifferent context.

Implicatures are therefore very different in nature from the two

other major kinds of inference, entailment and presupposition;

among other things, neither of these two can be denied without

producing anomaly.

Implicatures therefore belong squarely to the domain of prag-

matics, the study of how meanings are extracted from context.

Their existence was uncovered by the British philosopher PaulGrice in the 1960s, and Grice’s cooperative principle represents the

major attempt at explaining how speakers are successful in com-

municating such meanings. A more recent development claims that

all such inferences are the product of a principle of relevance.

See also: cooperative principle; entailment; pragmatics; presupposition;relevance

Further reading: Cutting 2002; Hurford and Heasley 1983; Levinson1983; Thomas 1995; Yule 1996.

COOPERATIVE PRINCIPLE

A fundamental principle governing conversational exchanges. In

the 1960s, the British philosopher Paul Grice undertook an exam-ination of the way people behave in conversation. His fundamental

conclusion was that conversational exchanges were governed by an

overarching principle, which he named the cooperative principle.Essentially, this principle holds that people in a conversation nor-

mally cooperate with one another, and, crucially, that they assume

that the others are cooperating. That is, when you say something,

COOPERATIVE PRINCIPLE

57

and another person makes a response, you assume that the

response is intended as a maximally cooperative one, and you

interpret it accordingly. Note that this term is not used in an ideo-

logical sense: participants in arguments, deliberate deception, lying,fiction, hypothesizing and making errors are still ‘cooperating’ in

the pragmatic sense.

It is this principle which is responsible for the existence of con-

versational implicatures, powerful inferences which are not logically

valid but which are derived from the assumption that the other

person is cooperating to a maximum extent.

Grice further decomposed his principle into a number of more

specific components, the maxims of conversation, such as ‘Makeyour contribution as informative as is required’, ‘Be relevant’, and

‘Do not say that for which you lack adequate evidence’. These

maxims are there to be broken, but when they are obviously broken

(flouted), the hearer can generate certain implicatures from that

fact.

See also: conversational implicature; relevance

Further reading: Blakemore 1992; Cutting 2003; Hurford and Heasley1983; Sperber and Wilson 1995; Thomas 1995; Yule 1996.

COORDINATE STRUCTURE

A grammatical structure consisting of two or more units of equal

rank joined by a connecting word. In English, a coordinate struc-ture most usually consists of some grammatical units connected by

a conjunction like ‘and’ or ‘or’. As a rule, the two units, or con-juncts, which are combined, or conjoined, must belong to the same

syntactic category. Here are some examples, in which the coordinate

structures are bracketed and so are the conjuncts they consist of:

[[Susie] and [her parents]] are coming (conjoined noun phrases);

Susie [[undressed] and [took a shower]] (conjoined verb phrases);

Hungarian is spoken [[in Hungary], [in much of Romania] and [in part

of Serbia]] (conjoined prepositional phrases); Does she drive [[well]

or [badly]]? (conjoined adverbs).

In certain circumstances, we can conjoin two units of different

categories: She polished the table [[lovingly] and [with great care]]

(adverb conjoined with prepositional phrase). But in most cases this

is not possible: *She smokes [[Marlboros] and [too much]] (noun

COORDINATE STRUCTURE

58

phrase wrongly conjoined with adverb phrase) (the asterisk marks

ungrammaticality).

A few coordinate structures exhibit more complex patterns, as in

Susie is [neither [Irish] nor [Welsh]]. And a few constructions thatlook broadly like coordinations have decidedly unusual structures,

such as [Janet prepared] and [Zelda served] [the sandwiches] (this is

called right-node raising), and [Esther ordered chicken Kashmir] and

[Larry, lamb rogan josh] (this is called gapping); such constructions

present formidable difficulties of analysis. Note that the bracketing

system adopted here is convenient for linear text, but expressing

syntax in a tree diagram is much clearer and more common.

See also: conjunction; tree

Further reading: Collins Cobuild 1990; Crystal 1996; Hurford 1994;Jeffries 2006.

COPULA

A specialized grammatical item, often a verb, which serves only to

express identity or class membership. The English copula is be,

and this verb has two main functions. First, as the verb in anequational sentence, it expresses identity and functions rather like

an equal sign in mathematics: The largest planet in our solar

system is Jupiter. Such a sentence can be readily reversed: Jupiter

is the largest planet in our solar system. Second, as the verb in an

ascriptive sentence, it ascribes some property to its subject, or, in

other words, it assigns its subject to membership in some class:

Susie is clever; Susie is sleepy; Susie is a woman with a red car.

Here certain properties are being ascribed to Susie (cleverness,sleepiness, having a red car), or, equivalently, Susie is being

assigned to the class of clever people, to the class of sleepy people,

or to the class of car-owners. Such sentences become unnatural or

worse when reversed: ??Clever is Susie; ??A woman with a red car is

Susie.

Of course, English be can also function as an auxiliary, and in

that case it is not serving as a copula.

See also: auxiliary; verb

Further reading: Hurford 1994; Jackson 2002.

COPULA

59

CREOLE

A language developed from a pidgin. A pidgin is a combination of

two or more languages which sometimes occurs in trade contact,multi-ethnic or refugee situations, where participants need a func-

tioning common language. It is usually reduced in complexity and

function and is learned as a second language by all speakers. Some-

times the pidgin becomes stable and established and comes to be

spoken as a mother-tongue by children: the language has then

become a creole, which quickly develops in complexity and is used

in all functional settings. The process of turning a pidgin into a

creole is creolization.Countless creoles have come into existence during the past few

centuries, often because of the activities of European colonists.

Speakers of English, French, Spanish, Portuguese and Dutch have

established colonies in Africa, Asia and the Americas, in areas

where the local languages were very different, and in many cases the

Europeans imported African slaves speaking any of dozens of

African languages. The Caribbean has been a particularly fertile

area for creoles, as Europeans and Africans (and to a lesser extentNative Americans) were forced to construct innumerable local pid-

gins, very many of which went on to be converted to creoles.

When a creole remains in contact with the prestige language

from which it was largely constructed, it may undergo significant

decreolization – adjustment towards that prestige standard – and

the result may be a creole continuum, a range of varieties from a

highly conservative version of the creole (the basilect) through

increasingly decreolized versions (the mesolects) to something moreor less identical to the prestige standard (the acrolect). This strati-

fication typically reflects the degree of education or a continuum

between rural and urban speakers.

The study of creoles was pioneered by the Trinidadian John

Thomas, the American Addison Van Name and the German Hugo

Schuchardt in the late nineteenth century, and the topic has never

since been really neglected, but it has prospered particularly since

the 1970s, and it is now regarded as a major area of investigation.Linguists studying contemporary language change have found

creolization to be a rich source of information, particularly from

the point of view of the construction of new grammatical systems.

The remarkable similarities in grammar among creoles all over the

world have led to the proposing of the bioprogram hypothesis. At

one time, there was a widespread belief that all creoles were

CREOLE

60

descended from a single ancestral creole by massive vocabulary

replacement (relexification), but this idea in its simple form is no

longer taken seriously. A variation on the notion is that syntactic

similarities in unrelated creoles can be ascribed to the similarities insocial setting and function of those situations, and these structures

are then relexified by the local influential language.

See also: bioprogram hypothesis; pidgin

Further reading: Chaudenson 2001; Holm 1988–9, 2000; Kouwenbergand Singler 2005; Romaine 1988; Sebba 1997.

CRITICAL DISCOURSE ANALYSIS

The analysis of texts within their social context. It is possible, of

course, to examine a text from a purely structural point of view: the

vocabulary and constructions it employs, the linguistic devices it

uses to relate one part to another, and so on. But the approach

called critical discourse analysis is rather different. This approach isprimarily interested in the social context in which a text is written.

(Critical discourse analysis is particularly associated with the work

of the British sociolinguist Norman Fairclough, and it has become

particularly influential in Europe and Australasia.)

Why was this text constructed at all? To whom is it addressed,

and why? Does the writer or speaker have concealed purposes, and,

if so, what are they? What hidden assumptions and biases underlie

the text? These are the sorts of questions pursued in critical dis-course analysis. The linguistic techniques involved in such analysis

are often called critical linguistics, and the educational policy of

teaching people to be alert to such matters is critical languageawareness.

A simple example is provided by headlines and stories in differ-

ent newspapers reporting the same story. Compare the following

possible headlines for the same real-world event: Police shoot

demonstrators; Demonstrators are shot; Shooting at demo; Demo

ends in violence; 2 dead at demo; Police make arrests as 2 die in demo

riot. A critical discourse analyst would consider the different ideol-

ogies encoded by these different linguistic forms, such as the impli-

cit assignment of blame and the shifting of emphasis.

See also: discourse; text

CRITICAL DISCOURSE ANALYSIS

61

Further reading: Carter 1997; Chilton 2003; Fairclough 1992, 1995,2001, 2003; Pennycook 2001; Wodak 1996; Wodak and Chilton 2005.

CRITICAL PERIOD HYPOTHESIS

The hypothesis that a first language can only be acquired during

the first few years of life. Young children learn perfectly any lan-

guage to which they are adequately exposed, and they do this

without explicit teaching. Few adults can perform the same feat. In

the 1960s the American neurologist Eric Lenneberg proposed an

explanation: we are born with a singular ability to learn languages,but this ability is ‘shut down’, probably by some genetic program-

ming, at around age thirteen, the cut-off age for first-language

acquisition.

Strong support for Lenneberg’s hypothesis comes from the

observation of feral children: children who, for some reason, have

been denied normal access to language in early life. In the eight-

eenth century, a young teenage French boy later named Victor was

discovered living wild. He had no language and failed to learnmuch after being taken into care. More recently, a French girl

known as Isabelle and an American girl known as Genie were pre-

vented by psychopathic parents from hearing any language. After

discovery and rescue, Isabelle, who was six, learned French rapidly,

and quickly caught up with other children of her age, but Genie,

nearly fourteen when discovered, never learned more than a mini-

mal amount of English, in spite of intensive therapy. An American

woman known as Chelsea was born nearly deaf, but was mis-diagnosed as mentally retarded. Only at age thirty-one was she

correctly diagnosed and given a hearing aid; she then began learn-

ing English but she too never made more than minimal progress.

See also: language acquisition; language instinct

Further reading: Aitchison 1989; Field 2003; Steinberg 1993.

DEAD LANGUAGE

A language which is no longer spoken. The term dead language is

applied to two quite different cases, which you should be careful to

distinguish.

CRITICAL PERIOD HYPOTHESIS

62

In the first case, a dead language is a language which has

disappeared as a mother tongue because its speakers abandoned

it in favour of some other language (or, sometimes, because its

speakers were all killed). This has happened countless times. Anumber of languages of ancient Anatolia and the Middle East

have disappeared in this way, including Hittite, Sumerian and

Akkadian. An unknown number of European languages dis-

appeared in favour of Latin during the Roman Empire, including

Etruscan in Italy, Gaulish in Gaul and Iberian in Spain. In Britain,

Cornish and Manx have vanished in modern times in favour of

English. And hundreds of indigenous languages in Australia and

the Americas have been abandoned in favour of English, Spanishand Portuguese.

A slight complication is that one or two of these truly dead lan-

guages, such as Cornish, have today a few speakers; these people

speak the dead language as a second language, having learned it

from books. Whether we continue to regard such a language as

dead is a matter of taste, but most linguists would probably regard

Cornish as dead, since it perhaps has no native speakers. A few

dead languages may continue to find some use as religious or lit-erary languages, but again, in the absence of native speakers, they

may be regarded as dead.

In the second case, the language in question never ceases to be

spoken as a mother tongue, but over a period of time it changes so

substantially that its later forms are so different from the earlier

form (and often from one another) that it no longer makes sense to

apply the same name. This has happened to Latin. Latin has never

ceased to be spoken in Portugal, Spain, France and Italy, but themodern forms of Latin are so different from the language of the

Romans, and from one another, that we no longer find it con-

venient to call them ‘Latin’. Hence we speak of French rather than

‘Paris Latin’, Catalan rather than ‘Barcelona Latin’, and so on.

See also: language death

Further reading: Crystal 2000; Holmes 1992; Romaine 1994.

DEEP STRUCTURE

An abstract representation of the structure of a sentence posited by

a linguist for analytical purposes. The concept of deep structure

DEEP STRUCTURE

63

was a central part of the first versions of transformational gram-

mar, introduced by Noam Chomsky in the 1950s. Chomsky’s

idea was that certain important generalizations about the struc-

tures of the sentences in a particular language were difficult tostate in terms of the surface structure of sentences, but that these

generalizations could be readily expressed in a theoretical fra-

mework in which sentences were assumed to have abstract

underlying forms which were very different from their surface

forms.

Chomsky defended his idea by appealing to examples like John is

eager to please and John is easy to please. On the surface, these may

appear to have identical structures, but, crucially, they receive verydifferent interpretations. Using the symbol NP for a noun phrase

which is in some sense ‘missing’ on the surface, we might therefore

posit deep structures for these two sentences of approximately the

following forms: John is [eager to please NP], but [NP to please

John] is easy. These representations now allow us to explain the

meaning of each sentence, but they must be modified by the action

of powerful rules called transformations in order to produce the

correct surface forms in each case.The conception of deep structure has changed substantially over

the years, as Chomsky and his colleagues have continued to modify

their ideas, and the several successive versions of transformational

grammar (and its successor Government-and-Binding Theory) have

involved very different views of deep structure, which has more

recently been called D-structure. Consequently, different textbookswill present significantly different versions of it, and the most recent

versions are often very abstract indeed. Most theories of grammarother than transformational grammar have declined to recognize a

concept of deep structure, preferring instead to work solely with

surface structures.

When deep structure was first introduced, there was some

enthusiasm for seeing it as representing a mental reality, some-

thing actually present within speakers’ brains. This idea has gra-

dually been abandoned, and today only a minority of linguists

would want to see D-structure as anything more than an analyticalconvenience.

See also: ambiguity; surface structure; transformational grammar

Further reading: Kroeger 2004; Lyons 1991; van Valin 2001.

DEEP STRUCTURE

64

DEFICIT HYPOTHESIS

The hypothesis that a group of speakers has an inadequate com-

mand of grammar and vocabulary to express complex ideas. In the1960s, the British educational theorist Basil Bernstein proposed

that a given language can be regarded as possessing two funda-

mentally different styles, or codes. A restricted code, in this view, has

a limited vocabulary and a limited range of grammatical construc-

tions; it is adequate for talking to people with very similar back-

grounds about everyday experiences, but it is highly inexplicit and

depends for success upon a large degree of shared experience. It is

too inexplicit and too limited to express complex and unfamiliarideas in a coherent manner. An elaborated code, in contrast, pos-

sesses a large vocabulary and a wide range of grammatical con-

structions, and it is entirely suitable for communicating complex

ideas, in a fully explicit manner, to people who do not share the

speaker’s background. Bernstein’s deficit hypothesis holds that,

while middle-class children have full control over both codes,

working-class children have access only to the restricted code.

Hence working-class children cannot communicate effectively in themanner expected in educational institutions, and so they cannot

hope to succeed in schools which are largely predicated on elabo-

rated code.

This hypothesis has generated a storm of discussion and debate.

Linguists, led by William Labov, have mostly been critical and dis-

missive of it. They defend instead the difference hypothesis – by

which working-class speech is merely different from middle-class

speech, and not inferior to it in expressiveness – and, hence, thatworking-class children in school are penalized only for being dif-

ferent, and not for being incompetent. The notion of deficit has

been debated also in relation to linguistic differences ascribed to

gender and ethnicity.

See also: standard language

Further reading: Emmitt et al. 2006; Holmes 1992; Stockwell 2002a.

DEICTIC CATEGORY

Any grammatical category which expresses distinctions pertaining

to the time and place of speaking or to the differing roles of

DEICTIC CATEGORY

65

participants. The word deixis means ‘pointing’, and deictic expres-

sions serve to anchor the speaker in relation to their surroundings

and other participants. Egocentric deixis refers to linguistic markers

whereby person, place and time are used in relation to the speaker(I, here, now). Clearly, person, tense and locative adverbials are the

main carriers of these three deictic dimensions.

Further deictic categories have been proposed that extend the

notion or try to adapt it to situations where speakers imagine other

viewpoints (as in narratives and fiction, for example). So textual,compositional and relational deixis have also been explored in rela-

tion to social marking, and delineating the different ‘voices’ in split

narratives. Cognitive linguists in particular have been interested inthe ways that the deictic centres expressed by consistent deictic

choices are maintained and shifted.

See also: cognitive linguistics; grammatical category

Further reading: Duchan et al. 1995; Green 1995; Levinson 1983; Lyons1968; Rauh 1983; Stockwell 2002b.

DENOTATION

The central meaning of a linguistic form, regarded as the set of

things it could possibly refer to. The study of meaning is a complex

affair, and several quite different kinds of meaning have to be

carefully distinguished before we can hope to make much progress.

For example, when you say The cat is scratching the sofa, youclearly have some particular, individual cat in mind, and the rela-

tion between the cat and that animal is one of reference. Now the

word cat itself cannot normally refer to any particular entity in this

way. However, one way of looking at the central meaning of cat is

to see this as consisting of all the cats in the (real or conceptual)

world – that is, as the totality of things to which the word cat might

reasonably be applied. This interpretation is called the denotation of

the word cat.Denotation is a difficult concept to work with, since concepts like

‘all the cats in the world’ are almost impossible to pin down.

Among ‘all the cats in the world’, should we include all those cats

which have not yet been born, and all those which died millions of

years ago? Nevertheless, denotation is often invoked in semantics,

and formal versions of semantics often try to formalize denotation

DENOTATION

66

as what is called extension: the extension of cat is the set (in the

formal mathematical sense) of all the entities in the universe of

discourse (the totality of things we can talk about) to which cat can

be applied.Denotation is most frequently contrasted with connotation, but it

has important similarities to sense, which is essentially a more

directly linguistic way of interpreting the same kind of meaning.

(And some writers have a habit of using denotation almost inter-

changeably with reference, but this is inappropriate.)

See also: connotation; reference; sense

Further reading: Frawley 1992; Hofmann 1993; Saeed 1997.

DEPENDENCY

A grammatical link between two (or more) different points in a

sentence. In a dependency, the presence, absence or form of an

item at one point in a sentence is directly linked to the presence,absence or form of a second item at a different point in the same

sentence. There are several types of dependency. When the pre-

sence of one item requires the presence or absence of another, we

have subcategorization (for example, the verb slap requires an

object noun phrase, while smile does not allow one). When the

presence of one item requires a particular form on a second item,

we have government (for example, the German preposition mit

‘with’ requires its object to take the dative case). When the formof one item requires a particular form on a second item, we have

agreement (for example, a plural subject requires a plural verb-

form).

All these are local dependencies, in which the two ends of the

dependency are not allowed to be separated by more than a cer-

tain amount. In an unbounded dependency (or extraction), there is

no limit on how far apart the two ends can be. The example

*Susie slapped is ungrammatical (marked by the asterisk), sincethe verb lacks the object it requires. Likewise, *Who did Susie slap

Louise? is ungrammatical, since who cannot be used unless there

is a suitable gap, or ‘hole’, elsewhere in the sentence. But Who did

Susie slap e? is fine, since who and the missing object (marked by

the symbol e) permit each other to occur. And the question

word and the gap can be arbitrarily far apart: Who did Archie say

DEPENDENCY

67

that Bill thought that Claire believed that Donna suspected Susie

slapped e?

See also: agreement; government; subcategorization

Further reading: Brown and Miller 1991; Greville 2006.

DERIVATION

Constructing new words by adding affixes to existing words. In

most languages, derivation is one of the principal ways of obtainingnew words from existing words, and its study is one of the major

branches of morphology.

The key point is to distinguish derivation from inflection. When

we add certain affixes to a lemma like write, producing forms like

writes, writing and written (and also, in a more complex manner,

wrote), we do not get any new words, but only grammatically dis-

tinct forms of the same word: this is inflection. You wouldn’t expect

to find different dictionary entries for all these forms: there wouldjust be the one entry for all of them, under write. However, other

affixes produce genuinely different words, such as rewrite, under-

write and writer, and these are examples of derivation. This time

you would expect to find separate dictionary entries for these words,

though a small dictionary might not bother with rewrite, since its

meaning is so obvious.

Like many languages, English is rich in both derivational prefixes

and derivational suffixes. Examples of the first are re-, anti-, syn-,counter-, non-, un-, trans-, pre- and mis-. Examples of the second

are -ness, -ity, -less, -wise, -ize, -dom, -ly (two different ones), -er and

-(at)ion. Multiple affixes are possible, though normally there are

strict rules governing the order in which affixes may be added.

Starting from happy, we can derive first unhappy and then unhappi-

ness. Starting with derive, we can obtain first derivation, then deri-

vational, and finally the very obscure technical term in linguistics

transderivational. Starting with exist, we can successively deriveexistent, existence, existential and existentialism. In every case, at

every stage, the result is a new word which deserves its own entry in

the dictionary.

In transformational grammar, the complete set of stages linking

the deep structure of a sentence to its surface structure is also called

derivation. Among the various theories of grammar, transformational

DERIVATION

68

grammar is distinguished by its claim that the syntactic structure of

a sentence is not a single tree, but rather a series of trees. The most

fundamental level of structure is the deep structure of the sentence,

and the most superficial, the surface structure. These two levels ofrepresentation are typically linked by a whole series of trees, each

one resulting from the application of a transformation to the

preceding one. The ordered series of trees which results is the

derivation of that particular sentence.

In this sense, the term derivation is also applied to the series of

stages involved in process-based theories of phonology like gen-erative phonology in converting an underlying form into a surface

form.

See also: affix; inflection; morphology; phonology; transformational gram-mar; word-formation

Further reading: Brown and Miller 1991; Epstein and Seely 2002;Katamba 1994.

DESCRIPTIVISM

The policy of describing languages as they are found to exist. A

prominent feature of traditional grammar is the frequent presence of

prescriptivism: identifying and recommending forms and usages

favoured by the analyst and condemning others not favoured by the

analyst. Excepting only in certain educational contexts, modern

linguists utterly reject prescriptivism, and their investigations arebased instead upon descriptivism. In a descriptivist approach, we

try to describe the facts of linguistic behaviour exactly as we find

them, and we refrain from making value judgements about the

speech of native speakers. Of course, our descriptions sometimes

include the observation that speakers themselves regard certain

usages as good or bad, but that is a very different thing from

expressing our own opinions.

Descriptivism is a central tenet of what we regard as a scientificapproach to the study of language: the very first requirement in any

scholarly investigation is to get the facts right. Prescriptivism, in

great contrast, is not a scientific approach. The strong opinions of

prescriptivists may be variously regarded as recommendations

about good style, as an aspect of social mores, as a consequence of

our educational system, or perhaps even as a matter of morality,

DESCRIPTIVISM

69

but they are not statements about actual behaviour, and hence they

are not scientific.

See also: prescriptivism

Further reading: Pinker 1994; Trask 1995.

DESIGN FEATURES

An informal list of the seemingly universal properties of human

languages. The idea of design features was introduced by theAmerican linguist Charles Hockett in 1960; both Hockett and

others have occasionally proposed revisions to the original list, and

several versions are in print. All contain the fundamental features

of arbitrariness, duality of patterning, displacement, open-endedness

and stimulus-freedom. Of these five, only the first is also normally

found in animal communication.

Sign languages present a few complications. They clearly do not

possess the secondary design feature of using the vocal-auditorychannel as their primary medium, and it is debatable whether they

possess duality of patterning.

See also: animal communication; arbitrariness; duality of patterning;displacement; open-endedness; stimulus-freedom

Further reading: Crystal 1997; Hockett 1960; Trask 1995.

DETERMINER

The part of speech which includes words like the and my. The Eng-

lish determiners are a smallish class of chiefly grammatical items

which have only a single function: they typically occur as the first

item in a noun phrase. Here is a simple test for determiners. Any

single word which can fit into the blank in the following frame toproduce a noun phrase is a determiner: – new book. Examples: the,

a, this, that, some, every, no, my, her, which. There are some further

determiners which can only fit into plural noun phrases, as in: –

new books. Examples: these, most, both, all, few, several. But be

careful here: certain words which are not determiners will also fit

into this second blank (entirely, attractive, other), but these items

DESIGN FEATURES

70

require entirely different syntactic structures to fit into this string of

words.

The two most highly grammatical determiners, the and a(n), are

called articles. The ones like my and her are traditionally calledpossessive pronouns, but grammatically they are determiners, not

pronouns.

Normally a noun phrase contains only one determiner. But cer-

tain noun phrases appear to contain two: all my children, both these

books. In such cases, the first item is often called a predeterminer. Anoun phrase which is headed by a singular uncountable noun or by

any plural noun need not have an overt determiner: French wine,

new books. Some linguists prefer to say that such noun phrasescontain a zero determiner. Some (not all) determiners have mean-

ings involving quantity, such as many, several and all. These are

called quantifiers, and some linguists prefer to separate the quanti-

fiers into a separate part of speech from determiners, but there is

little or no grammatical justification for this.

See also: noun phrase

Further reading: Collins Cobuild 1990; Greenbaum and Quirk 1990;Jackson 2002.

DIACHRONY

The time dimension in language. It was the Swiss linguist

Ferdinand de Saussure, in the early twentieth century, who first

emphasized the fundamental difference between synchrony and dia-chrony in the study of language. In a diachronic approach, we look

at how a language has changed over some period of time. Most

work in historical linguistics is diachronic in nature, but not all ofit: a linguist might well be interested in constructing a purely

synchronic description of, say, the Old English of King Alfred’s

day or the Latin of Caesar’s day, without considering how the lan-

guage had developed from an earlier form or what happened to it

later.

See also: historical linguistics; language change; Saussurean paradox;synchrony

Further reading: Fennell 2000; McMahon 1994.

DIACHRONY

71

DIALECT

A more or less identifiable regional or social variety of a language.

Every language that is spoken over any significant area is spoken insomewhat different forms in different places; these are its regionaldialects. Moreover, even in a single community, the language may

be spoken differently by members of different social groups; these

different forms are social dialects or sociolects. For example, the

English of London is noticeably different from the English of Bir-

mingham, Liverpool, Glasgow, New York, New Orleans or Sydney,

and even within London stockbrokers tend not speak like motor

mechanics.It is important to realize that everybody speaks some dialect or

other; it is not possible to speak a language without using some

dialect. (In fact, you might say that dialects exist but language is

merely an abstract term for a collection of dialects.) Informally, we

often reserve the label dialect for a speech variety which is notice-

ably different from our own, or which is lacking in prestige, but this

is not the way the term is used in linguistics. In British usage, the

term dialect includes only features of grammar and vocabulary,while features of pronunciation are treated under the quite different

heading of accent. In American usage, an accent is usually con-

sidered to be just one part of a dialect.

The study of regional dialects, known as dialect geography or

dialectology, has been a major part of linguistics since the late

nineteenth century; there have been many studies of regional var-

iation, often resulting in the publication of dialect atlases contain-

ing a series of dialect maps, each showing the variation in respect ofa single feature (the lines which separate features on the map are

known as isoglosses). Dialectology has had a distinctively historical

dimension, with researchers often using historical documents, place

names and archaeological evidence to investigate dialectal change

and diffusion, and also using dialect evidence for historical research

in return. In contrast, social dialects (in the form of urban dia-lectology) have only been seriously studied since the 1960s; the

pioneering work here was done by the American linguist WilliamLabov. While this sociolinguistic work has also provided insight into

long-term language change, its main focus has been synchronic, and

it has tended to regard dialect as simply one type of social language

variation.

A standard language is a rather special dialect of some language,

one which has been codified (standardized in dictionaries and

DIALECT

72

published grammars, and used in the education system and for

prestigious texts) and elaborated for use in a wide variety of

domains.

Linguistically unsophisticated people sometimes apply the termdialect to a regional language of low prestige, but the term is never

so used in linguistics, in which a dialect is always a variety of a

language which has other varieties.

See also: accent; onomastics; standard language

Further reading: Auer et al. 2005; Chambers and Trudgill 1998; Crystal1997; Foulkes and Docherty 1999; Llamas et al. 2007.

DIGLOSSIA

Marked specialization of function between two language varieties in

a single speech community. It is by no means rare for two or more

distinct languages or language varieties to be used side by side within

a single community, with or without a high degree of bilingualism.For example, many citizens of Spain routinely switch between

Basque, Catalan or Galician on the one hand and Castilian Spanish

on the other, depending on the circumstances; German-speakers in

Germany, Switzerland and Austria likewise switch between stan-

dard German and their own local varieties of German, which are

often not comprehensible to other speakers; all of English, Malay,

Cantonese Chinese and Tamil are widely spoken in Singapore,

though very few people there can speak all four. In most such cases,people naturally prefer to speak their own mother tongue whenever

they can, and they switch to another language or variety only when

they have to. In a few communities, however, something very dif-

ferent happens: the languages or language varieties come to be

perceived as having different functions in the community as a

whole, and hence each variety is used more or less exclusively for

those functions in which it is deemed appropriate. Normally only

two language varieties are involved in such a case, and the result isdiglossia.

Diglossia was first identified as a distinctive phenomenon by the

American linguist Charles Ferguson in the 1960s. Ferguson’s initial

characterization has since been modified very slightly, but the

characteristics of a diglossic society are essentially the ones he

identified. There is a clear difference in prestige between the two

DIGLOSSIA

73

language varieties: one, called High (or H), enjoys great prestige,

while the other, called Low (or L), enjoys little or no prestige; in

extreme cases, speakers may deny the very existence of L. In all

cases, L is the mother tongue of all or most speakers, while H islearned only through formal education. Speakers of limited educa-

tion may have a very inadequate command of H, and they may

even have trouble understanding it.

The specialization of function is highly predictable from one

diglossic society to another. The L variety is used for ordinary

conversation and for the more popular types of entertainment (such

as soap operas and commentary on sports events); it is rarely writ-

ten, and may well lack a recognized written form. However, it maybe used in comic strips, in captions to political cartoons, in scurri-

lous publications, and perhaps in personal letters. The H variety is

used in newspapers and most other publications, for all serious lit-

erature, for university lectures, for news broadcasts and other

formal types of radio and television broadcasts, and (usually) for

religious purposes.

So well entrenched is this perceived specialization that using the

‘wrong’ variety for a particular purpose will be seen as comical oroffensive: even speakers with a minimal command of H prefer to

hear H when H is appropriate, since the use here of L, which they

can understand perfectly well, is felt to be undignified or worse. In

one famous incident, there were riots in the streets of Athens in

1901 upon the publication of a translation of the New Testament

into the Low variety of Greek: thousands of Greeks were enraged

by the use of L in such a solemn religious context, and they insisted

upon an H version which many of them could not understand.Among the diglossic societies identified several decades ago

were:

� Greece – H = Katharevusa, a kind of reconstructed classical

Greek, L = Dhimotikı, ordinary spoken Greek

� German Switzerland – H = standard German, L = Swiss

German

� the Arab countries – H = the classical Arabic of the Koran, L =ordinary spoken Arabic

� Paraguay – H = Spanish, L = Guaranı, the mother tongue of

most of the population and a Native American language.

Changing political circumstances have brought diglossia to an

end in Greece, and there are signs that the Arab countries may be

DIGLOSSIA

74

going the same way (though diglossia is far from dead here);

Switzerland and Paraguay continue much as before. Other

instances of diglossia have arisen in the past and have often

proved highly stable. A good example is medieval Europe, inwhich Latin (H) was used for all serious purposes, while the

innumerable local vernaculars (L) remained the everyday speech of

the entire population, most of whom knew nothing of Latin. This

state of affairs persisted for centuries before Latin finally gave way

to the new national languages like French, Spanish, Italian and

German, which had previously been regarded as unfit for serious

purposes.

See also: bilingualism

Further reading: Ferguson 1959; Holmes 1992; Stockwell 2002a; Ward-haugh 2005.

DIRECT OBJECT

The grammatical status of a noun phrase which occurs inside a verb

phrase and which is the second obligatory argument of a transitive

verb. For example, Susie threw the ball features a transitive verb

threw with the direct object the ball. The direct object here functionsas the patient or beneficiary upon which the verb acts. Intransitive

verbs (die, go, fell) cannot take direct objects (*He died her, *She

went her mother’s,* I fell the hill), except in creative or deviant cir-

cumstances. Direct objects are marked for case in case languages as

accusative in languages like Latin or Russian (or the English case-

remnant accusative me in He hit me rather than the nominative *He

hit I), and are marked in the absolutive case in ergative languages

like Basque.Direct objects have the property of being able to become the

subject of the passive form of the sentence (The ball was thrown by

Suzie; compare *She was died, *The hill was fallen by me). In

Government-and Binding theory, indirect objects are not recognized

so direct objects are simply classified as objects.

See also: case; transitivity

Further reading: Blake 2001; Jackson 2002.

DIRECT OBJECT

75

DISCOURSE

In sociolinguistics and critical discourse analysis, discourse is any

connected piece of speech or writing in its social context (this lastcircumstance sets discourse apart from text). In applied linguistics,

discourse refers more narrowly to the interactive and commu-

nicative dimension of language, and involves conversation analysis,

semiotics, and the dynamic processes of text production and under-

standing. While usage varies, we most commonly apply the label

discourse analysis to an approach which is based heavily upon tradi-

tional grammatical concepts, conversation analysis to an empirical

approach which rejects traditional concepts and seeks to extractpatterns from data, and text linguistics to the study of large units of

language each of which has a definable communicative function.

See also: cohesion; critical discourse analysis; discourse analysis

Further reading: Allen and Guy 1974; Carter 1997; Coulthard 1985;Cutting 2002; Mills 2004; Nofsinger 1991; Renkema 2004; Schiffrin1994; Sinclair and Carter 2004; Sinclair and Coulthard 1975.

DISCOURSE ANALYSIS

An approach to the study of discourse which is based upon tradi-

tional grammatical concepts and terms. In principle, we might apply

the label discourse analysis, or DA, to any kind of investigation of

the structure of discourse, but in practice the label is most commonlyreserved for an approach based upon familiar grammatical con-

cepts. That is, a proponent of discourse analysis comes to the ana-

lytical task with a complete battery of grammatical concepts and

terms of the sort familiar to any student of grammar and attempts

to see how these concepts are involved in structuring discourses. To

put it another way, DA is an attempt to extend our highly success-

ful analysis of sentence structure to units larger than the sentence.

Though there is considerable variation in practice, DA oftenbegins by trying to identify minimal units of discourse and then by

looking for rules governing how these minimal units can be strung

together in sequence to produce well-formed discourses, much as

smaller syntactic units are combined into sentences according to

the rules of syntax. DA thus contrasts strongly with the alternative

approach known as conversation analysis. Proponents of DA complain

DISCOURSE

76

that conversation analysis is hugely inexplicit and ad hoc, and

lacking in any identifiable underpinning, while the practitioners of

the other approach in turn accuse DA of being excessively a priori

and of paying too little attention to real texts, as opposed todeliberately constructed ones.

DA has been prominent since the 1970s; it is particularly important

in Germany and the Netherlands, where it is often almost indis-

tinguishable from text linguistics, but it has also been pursued with

some vigour in the English-speaking countries. More recently, dis-

course analysis has been invigorated by the use of corpus linguistics andthe availability of large corpora of spoken discourse transcriptions.

See also: computational linguistics; conversation analysis; text linguistics

Further reading: Brown and Yule 1983; Carter 1997; Eggins and Slade2004; Fairclough 1995, 2003; Levinson 1983; McCarthy 1991; Renkema2004; Wooffitt 2005.

DISPLACEMENT

The ability to speak about things other than the here and now.

With just a single known exception (see below), every signal used

by a non-human creature to communicate pertains wholly and

directly to the immediate time and place of signalling. No non-

human signal, with the marginal exception of scent markings left to

define territory or to provide a trail, ever refers to the past or the

future, to hypothetical or counterfactual states of affairs, or toanything not directly perceptible to the creature signalling. To put

this more picturesquely, mice do not swap stories about their close

encounters with cats, nor do bears soberly discuss the severity of

the coming winter; rabbits do not engage in heated discussions

about what might lie on the far side of the hill, nor do geese draw

up plans for their next migration.

Human language is utterly different. We have not the slightest

difficulty in talking about last night’s football game, or our ownchildhood, or the behaviour of dinosaurs which lived over 100 mil-

lion years ago; with equal ease, we can discuss political events in

Peru or the atmosphere of the planet Neptune. And, of course, we

can discuss what might have happened if the South had won the

American Civil War, and we can produce fables and fantasies

involving hobbits, dragons, talking animals and intergalactic wars.

DISPLACEMENT

77

All this is displacement: the ability to talk about things other than

what we can see, hear, feel and smell at the moment.

This displacement, which we take utterly for granted, is one of

the most momentous differences between human languages and thesignalling systems of all other species. They can’t do it at all; we do

it almost every time we open our mouths.

There is just one striking exception. A honeybee scout which has

discovered a source of nectar returns to its hive and performs a

dance, watched by the other bees. This bee dance tells the watching

bees what direction the nectar lies in, how far away it is, and how

much nectar there is. And this is displacement: the dancing bee is

passing on information about a site which it visited some time agoand which it now cannot see, and the watching bees respond by

flying off to locate the nectar. Startling though it is, the bee dance

is, so far at least, absolutely unique in the non-human world: no

other creatures, not even apes, can communicate anything of the

sort, and even the bee dance is severely limited in its expressive

powers: it cannot cope with the slightest novelty.

The importance of displacement was first pointed out by the

American linguist Charles Hockett in 1960.

See also: design features; open-endedness; stimulus-freedom


DISTINCTIVE FEATURE

Any one of a number of minimal phonological elements of which

speech sounds or phonemes are composed. For decades after the

phoneme concept was introduced into linguistics, linguists tended to

assume that each phoneme was an independent unit which could

not be analysed into any smaller units. But this view, while profit-

able, ran into serious difficulties in some respects. Most obviously, it

provided no basis for recognizing the natural classes of phonemes

which often need to be singled out in describing languages.Eventually, therefore, phonologists ceased to regard phonemes as

indivisible, and to treat them instead as bundles, or matrices, of

smaller components. These smaller components are distinctivefeatures, or features for short. In most cases, the features invoked

are binary, meaning that a feature can only be either present

(marked by [+]) or absent (marked by [–]).

DISTINCTIVE FEATURE

78

For example, all sounds which are produced with voicing, such as

[a m z d], carry the feature specification [+ voice], while voiceless

sounds, such as [p f h], are [– voice]. Similarly, sounds produced

with the velum lowered, like [m n a], are [+ nasal], while all othersare [– nasal]. A suitable set of such features, typically around fifteen

or so, is adequate both to distinguish every phoneme in a language

from every other, and also to characterize the required natural

classes: for example, the set /p t k/, which is a natural class in

English, might be singled out as [– continuant, – friction, – voice],

the class of voiceless frictionless stops.

The set of feature specifications identifying a particular segment

or class was long regarded as a mere unordered collection of itemswith no internal structure, but, since the 1980s, there has been a

marked tendency to assign to each set of features a kind of hier-

archical structure, in which some features are treated as subordinate

to other features; this approach is called feature geometry.The idea of distinctive features was first developed by the Eur-

opean linguists of the Prague School in the 1930s, first by the

Russian Nikolai Trubetzkoy, but most prominently by Roman

Jakobson. Jakobson later emigrated to the USA, and there, in col-laboration with the Swede Gunnar Fant and the American Morris

Halle, put together the first complete theory of distinctive features,

in the 1950s. This first effort was formulated in terms of acousticfeatures reflecting things that could be seen in a sound spectrogram,

but in the 1960s Halle and Noam Chomsky proposed a very dif-

ferent set of articulatory features, based chiefly on the activities of

the speech organs. Features of this sort have predominated in

phonology ever since, though phonologists continue to proposemodifications to the system even today.

See also: natural class; phoneme

Further reading: Ball and Rahilly 1999; Clark and Yallop 1995; Collinsand Mees 2003; Giegerich 1992; Hawkins 1984; Katamba 1989.

DISTRIBUTION

The set of positions in which a given linguistic element or form can

appear in a language. The notion of distribution is a central feature

of the approach to language study called structuralism, and it was

outstandingly important in the version called American structuralism.

DISTRIBUTION

79

Distribution is a simple notion. Any given linguistic element

which is present in a language, whether a speech sound, a phoneme,

a morpheme, a word, or whatever, can occur in certain positions

but not in other positions. A statement of its possible positions isits distribution, and this distribution is usually an important fact

about its place in the language.

For example, distribution is important in identifying parts of

speech. In English, any word which can occur in the slot in This – is

nice must be a noun, because English allows only nouns to occur in

this position. And larger syntactic categories can be partly identi-

fied in the same way: anything that can occur in the slot in – is nice

must be a noun phrase.But distribution is perhaps most prominent in phonology. Con-

sider the English labiodental fricatives [f] and [v]. Simplifying

slightly, in Old English, the sound [v] could only occur between

vowels, while [f] could never appear between vowels. Hence Old

English allowed words like [fi:f] ‘five’, [fæ:t] ‘fat’, [livian] ‘live’

(verb), and [ov

e

r] ‘over’, but no words like *[væ:t] or *[of e

r]. We say

that, in Old English, [f] and [v] were in complementary distribution,meaning that there was no position in which both could occur.Since the two sounds are phonetically similar, we can therefore

assign both to a single phoneme, usually represented as /f/. Indeed,

the Old English spellings of the four words were fif, fatt, lifian and

ofer, reflecting the fact that only one phoneme existed.

In modern English, however, the distribution of these two sounds

is very different: they can both occur in the same positions to make

different words. We thus have minimal pairs like fat and vat, fine and

vine, rifle and rival, and strife and strive. We therefore say that [f] and[v] are in contrastive distribution, and they must now be assigned to

separate phonemes, /f/ and /v/, just as the modern spelling suggests.

See also: phoneme; phonotactics

Further reading: Giegerich 1992.

DUALITY OF PATTERNING

A type of structure in which a small number of meaningless units are

combined to produce a large number of meaningful units. Non-

human creatures have signalling systems based upon ‘one sound,

one meaning’, and hence they can express only a tiny number of


80

meanings. Since we can scarcely produce more than 100 distin-

guishable speech sounds, and if our languages worked in the same

way, it would follow that we would not be able to produce more

than 100 distinct units of meaning.But human languages are organized differently. Each spoken

language possesses a small number of basic speech sounds: its pho-

nemes. The number varies from a minimum of ten to a maximum of

around 100, with the average being around thirty. Crucially, these

phonemes are themselves meaningless, but they can be combined

into sequences which are meaningful.

English has around forty phonemes (the precise number depend-

ing on the accent). Among these are /p/, /t/, /k/, and /æ/ (as in cat).Even just these four phonemes can be combined variously to pro-

duce a large number of words with very different meanings: /æt/ at,

/ækt/ act, /kæt/ cat, /pæt/ pat, /tæp/ tap, /kæp/ cap, /pæp/ pap, /pækt/

pact or packed, /tækt/ tact or tacked, /kæpt/ capped, /tæpt/ tapped,

and so on. Adding one more phoneme, the diphthong /eI/, we can

now get /peI/ pay, /keI/ Kay, /teIp/ tape, /keIk/, cake, /teIk/ take,

/keIp/ cape, /teIpt/ taped, and so on. With forty-odd phonemes,

English can produce a huge number of one-syllable words, but ofcourse English also has words that are several syllables long.

Duality thus allows a language to form many tens of thousands

of different words, all of which can be produced by a vocal tract

which can produce no more than a few dozens of distinguishable

speech sounds. Duality is therefore of crucial importance in facil-

itating the existence of spoken languages. Together with the gram-

matical property of recursion, duality allows human languages the

ability to produce an infinite number of utterances, all with differ-ent meanings, and hence makes open-endedness possible.

Sign languages perhaps lack duality; this has been much debated.

It is not in doubt that the signs of sign languages can be decom-

posed into smaller meaningless elements, but it is not clear at pre-

sent that these smaller elements form a system analogous to a set of

phonemes in a spoken language.

The importance of duality was first pointed out by the French

linguist Andre Martinet in the 1950s (he called it double articula-tion) and by the American linguist Charles Hockett in 1960; it was

Hockett who coined the term duality.

See also: animal communication; design features; phoneme; phonotactics



81

DYSLEXIA

A certain disability affecting reading and writing. Strictly, dyslexia(often informally called word blindness) is a disability with reading,

while the related disability with writing is dysgraphia, but the two

very commonly occur together. (The terms alexia and agraphia are

often preferred in North America.) A dyslexic has difficulty in per-

ceiving a printed page: the words on the page often appear dis-

torted, as though viewed through a misshapen lens; both the order

and the shapes of the letters may be perceived wrongly or appear to‘jump’. Similar problems affect writing: letters may be put down in

the wrong order or turned upside-down or backwards. Both in

reading and in writing, mirror-image letters like b, d, g and q may

be confused. In more severe cases, called deep dyslexia, words may

be confused with totally unrelated words, even those of very differ-

ent appearance, which are somehow similar in sound, meaning or

grammatical class: for example, saw with was, dinner with food, rib

with ride, bun with cake, saucer with sausage, for with and.As is usual with linguistic disfluencies such as stammering (stut-

tering in the USA), individuals vary significantly in the particular

symptoms they exhibit, and treatment or management varies cor-

respondingly. Also, as with other disabilities, dyslexia and dysgra-

phia may be present from early childhood in children showing no

sign of brain damage, or they may be acquired in adulthood as a

result of brain damage; the two cases are called developmental dys-lexia and acquired dyslexia, respectively. The former belief thatdyslexia could be traced to a single uniform cause is now known to

be false: dyslexia can in fact result from any of a number of differ-

ent causes, and many specialists suspect that any given case of dys-

lexia probably results from the interaction of several distinct

factors. Though linguistic disfluencies have in the past been taken

as signs of low intelligence, research has shown this is not the case.

See also: aphasia; language disability

Further reading: British Dyslexia Association 1996; Crystal 1997; Malm-kjær 2004.

ELICITATION TECHNIQUES

In most branches of applied linguistics such as sociolinguistics, psy-

cholinguistics and other sub-disciplines which are based on empiri-

DYSLEXIA

82

cal data rather than intuition, a key problem for the fieldwork

researcher is obtaining naturalistic data from language users. Where

anthropological linguists are investigating newly discovered lan-

guages, there is often no alternative but to learn the language byspending an extended period of time living in the speech commu-

nity. This sort of participant observation produces an intimate

encounter with the language under investigation, but it also runs

the risk of falling foul of the observer’s paradox (that the data will

be altered by the presence of the observer). Even in researching

familiar and well-documented languages like English, socio-

linguistic researchers have developed a large range of ingenious eli-citation techniques designed to gain access to natural language datawithout raising the awareness of the informants.

Techniques include asking informants emotionally involving ques-

tions (the ‘danger of death’ scenario) to produce unselfconscious

narratives; starting an interview session with casual conversation and

gradually increasing awareness through a formal setting, a reading

passage, and a list of minimal pairs; routinely ignoring the first few

minutes of recorded data so that informants become relaxed with the

recording equipment; surreptitious recording in a public place (seealso ethics); using a sense relation network – a diagram that links

semantically related words to elicit dialect terms; and many others.

Similar methods have to be used when the researcher is interested

in the intuitions, attitudes and opinions of the informants. Even

theoretical linguists are on safer ground when they base their

acceptability judgements for well-formedness on informants’ opi-

nions rather than exclusively on their own intuitions. In socio-

linguistics, a matched guise test is used in which a recording of thesame person speaking in different accents is played to informants,

in order to evaluate the informants’ attitudes to the accent as a

token for their personality and values. Certain accents – typically

urban and working-class – are evaluated negatively in terms of

educational level, wealth, trustworthiness and even good looks.

See also: ethics; sociolinguistics

Further reading: Llamas et al. 2007.

ELLIPSIS

The omission from a sentence or an utterance of material which

is logically necessary but which is recoverable from the context.

ELLIPSIS

83

Traditional grammarians have for centuries applied the term ellipsisto a wide range of phenomena in which some part of a sentence or

an utterance appears to be ‘missing’ or ‘understood’. Con-

temporary linguists, however, generally prefer to use the term morenarrowly to denote only the omission of material which can be

unambiguously recovered from the context.

So, for example, consider the following exchange. Mike: Where’s

Susie? Sarah: In the library. Here, Sarah’s response is unambigu-

ously interpretable in the context as meaning Susie is in the library,

and hence we speak of the ellipsis of the material Susie is. But now

consider another exchange. Mike: Here’s the book I promised you.

Sarah: Thanks. This time Sarah’s response cannot be specificallyidentified as a reduced form of any particular longer utterance, and

so we would probably not speak of ellipsis here.

Various particular types of ellipsis can occur within a single

utterance, and these often have individual names. Example: Susie

wants me to come to Greece with her, but I can’t. Here the missing

continuation is clearly come to Greece with her, and this construc-

tion is called VP-deletion. Another example: Somebody wants me

out, and I know who. The missing material is wants me out, and thisconstruction is sluicing.

Ellipsis must be carefully distinguished from elision, which is the

removal of sounds, as when we pronounce fish and chips as fish ’n’

chips, or when a British speaker pronounces library as libry. Sounds

are elided; words are ellipted.

See also: cloze; syntax

Further reading: Jackson 2002.

ENTAILMENT

A particular type of inference. If I say to you Booth assassinated

Lincoln, then, assuming my statement is true, you may safely draw

certain conclusions, including Lincoln is dead and Booth killed

somebody. These are among the several entailments of my original

sentence.

We say that statement P entails statement Q whenever the fol-

lowing inference holds: if P is true, then Q must also be true. Note

that any entailment of P never contains more information than P,

and in fact it usually contains less.

ENTAILMENT

84

Normally, if P entails Q, then Q does not entail P (check this

with the examples). But it is possible for P and Q to entail each

other, and in this case we are looking at paraphrases. Examples:

Sally sold a car to Mike; Mike bought a car from Sally.An entailment differs in several respects from a presupposition or

a conversational implicature. For one thing, an entailment is destroyed

by negation: the sentence Booth did not assassinate Lincoln does not

entail either Lincoln is dead or Booth killed somebody (presupposi-

tions are different here). For another, an entailment is totally inde-

pendent of any context (conversational implicatures are different in

this respect).

See also: conversational implicature; presupposition

Further reading: Hurford and Heasley 1983; Levinson 1983; Verschue-ren 1999.

ETHICS

The ethics of data collection have become particularly important in

those branches of linguistics which are based on eliciting natural-

language examples from non-professional informants. It was oncethought acceptable to record and collect examples of language from

people, and even live amongst them as a friend or visitor, without a

thought given to the cultural impact of the fieldwork and without

telling them – even after the research was completed – what the

researcher was doing at the time. Since the 1980s, especially in

sociolinguistics and anthropological linguistics, there has been a

greater emphasis on gaining the consent of speech communities for

the research. Where this is not possible because the observer’sparadox would affect the data collected, then permission and dis-

closure is effected after the study is finished. There are particular

problems involved in gaining access to vernacular speech and inti-

mate conversations, and, for different reasons, business and legal

discourse. Covert recording, even in a public place, would be largely

regarded as illicit. Many universities now have ethics committees (anextension of their role in biomedical research) which determine the

ethical standards of proposed research projects.

See also: elicitation techniques; observer’s paradox; sociolinguistics

Further reading: Cameron 1995; Llamas et al. 2007.

ETHICS

85

ETHNICITY

The relation between language use and ethnic background, espe-

cially in a mixed community, has become an important area ofstudy in linguistics. In modern urban societies, it is now common-

place to find speakers of a variety of ethnic backgrounds living

together and interacting in various ways. Examples include Anglos,

Latinos and black people in western American cities and speakers

of Anglo-Saxon, Caribbean and Asian origin in British cities.

Especially since the 1980s, some linguists have begun turning their

attention to the links between language and ethnic background.

In some cases, of course, some speakers are bilingual (for exam-ple, Spanish–English bilinguals in the USA and Punjabi–English

bilinguals in Britain). It is of interest to see how and in what cir-

cumstances such speakers change back and forth between their

languages (this is code-switching) and also to find out what influ-

ence each language has on the other. The British sociolinguist Ben

Rampton has identified the phenomenon of crossing, in which a

member of one ethnic group deliberately adopts the language or

usage of another ethnic group for specific social purposes. Linguistsare interested in determining the extent to which a given language

or variety serves as a badge of identity for a particular ethnic

group, and how that group’s variety of a language differs from

varieties of the same language used by other speakers.

The study of language and ethnicity has thus become a major

strand in sociolinguistics. Specifically, the study of language and

ethnicity is sometimes called ethnolinguistics.

See also: Black English; code-switching; identity

Further reading: Edwards 1994; Fishman 1999; Fought 2007; Harris2006; Rampton 1995; Saville-Troike 2003; Stockwell 2002a; Wardhaugh1987, 2005.

ETHNOGRAPHY OF COMMUNICATION

The study of the norms of communication in a speech community,

including verbal, non-verbal and social factors. Anthropological

linguists have long stressed the importance of examining commu-

nicative behaviour in the context of a culture, though the term ethno-graphy of communication itself was not coined until the 1970s, by

ETHNICITY

86

the American anthropological linguist Dell Hymes. Investigators

have looked at a broad range of variables in a number of speech

communities: loudness of voice, pitch of voice, distance between

speakers, expressions and postures, eye contact, terms of address,rules for initiating conversations, and many others.

For example, Mexicans in conversation prefer to stand much

closer together than do Americans, which can lead to comical

results when a Mexican is talking to an American. English-speakers

who meet each other but have nothing in particular to say will

begin talking about the weather, since silence is considered unac-

ceptable; but Chinese-speakers in the same position may choose to

remain silent without giving offence. Hungarian speakers offer lessback-channel noises (yes, mmm, uh-huh) than British speakers

require. In Japanese and in Javanese, even the simplest utterance

may assume any of a number of very different forms, depending on

the relative status of the speaker, the addressee and the person

being talked about, and also on the circumstances in which the

conversation takes place. Gesticulations accompanying speech

differ from one extreme (Italians) to the other (Swedes), and vary

enormously in form. Such gestures can contribute to the pragmaticsnot only of emphasis but also of denotation.

See also: non-verbal communication; paralanguage

Further reading: Bonvillain 1993; Duranti 1997; Foley 1997; Schiffrin1994.

ETYMOLOGY

The origin and history of a particular word, or the branch of lin-

guistics that studies this. Every language has a vocabulary contain-

ing many thousands of words, and every one of those words has its

own particular origin and history, that is its own etymology. Some-

times the word in question has been in the language as long as the

language has existed; in this case, the word has simply been inher-ited from some ancestral language. For example, English three

derives from Old English threo, which in turn derives from Proto-

Germanic *thrijiz, which in turn derives from Proto-Indo-European

*treyes, which in turn probably derives from the word for ‘three’ in

the unknown ancestor of Proto-Indo-European (the asterisk

denotes a reconstructed form). The same Germanic source yields

ETYMOLOGY

87

Gothic threis, Danish and Swedish tre, Dutch drie, German drei,

and so on. The same Indo-European source yields Latin tres, itself

the ancestor of Spanish tres, Italian tre, French trois, Romanian

trei, and so on. Also from the same Indo-European source comeSanskrit trayas, Russian tri, Polish trzy, Lithuanian trys, Albanian

tre, Irish trı, Welsh tri, and, less obviously, Persian se and Arme-

nian erek. The last two may look dubious, but generations of

patient and careful etymological work have established the histories

of all these words beyond dispute.

In other cases the word in question has not been inherited from

an ancestral language, but has instead entered the language at some

point in time; it may simply have been ‘borrowed’ (copied) from aneighbouring language, or it may have been coined by speakers

using the resources of the language. In these cases etymologists will

be deeply interested in finding out, as far as they can, just when the

word was first used, where and by whom, and in what sense. With-

out such evidence, etymology can be little more than guesswork.

Many thousands of English words have been (directly or indir-

ectly) borrowed from other languages in which they already existed:

angel from ancient Greek, lettuce from Latin, knife from Old Norse,face from Norman French, cigarette from modern French, skunk

from Massachusett, shampoo from Hindi, brandy from Dutch,

mosquito from Spanish, poodle from German, umbrella from Ita-

lian, alcohol from Arabic, ski from Norwegian, yogurt from Turk-

ish, ukulele from Hawaiian, whisky from Scots Gaelic, kayak from

an Inuit language, kangaroo from the Guugu-Yimidhirr language of

Australia, and so on. In such cases, etymologists may be interested

in going further, and in tracing the history of the word within theforeign language from which it is borrowed.

Many other English words have been constructed by English-

speakers in any of a large number of ways. Such words as ginger-

bread, paperback, scarecrow, spaghetti western, strip-tease, baby-sit,

laptop, word processor, underfunding, bewitch, megastar, non-

magnetic, miniskirt, edit, deli, flu, smog, burger, love-in, laser, giggle,

bang, scrooge, quixotic, gothic and malapropism have all been

coined in English by one means or another – some of them cen-turies ago, others very recently – and etymologists are interested in

all of them.

See also: loan word; onomastics

Further reading: Crowley 1997; Hock and Joseph 1996; Trask 1996.

ETYMOLOGY

88

EUPHEMISM

The use of a word or phrase as a synonym for another word which

is avoided because of its taboo status or because of its negativepolitical or ideological connotations. Famous examples of euphe-

mism include the huge number of terms for the lavatory (itself a

word that focuses on washing rather than defecating): w.c., toilet,

restroom, loo (from the French lieu ‘place’), netty (from the French

nettoyer ‘to wash’ or possibly the Italian gabinetto ‘cabin’, i.e.,

toilet), little boys/girls room, cloakroom, ensuite, convenience, and

euphemistic phrases such as powder my nose, pay a visit, freshen up

and many many more. Sex, death, race, drug abuse and the jargonof intimate groups are other areas of overlexicalized euphemism.

Political and military euphemism has been a particularly pro-

ductive area of lexicalization as governments have become more

democratically accountable and sensitive to adverse public opinion

of their wars: the USA–Vietnam war produced collateral damage

(civilian casualties); the Balkan war turned genocide into ethnic

cleansing; the USA–Iraq war produced surgical strikes (supposedly

precise bombing); the Israeli–Hezbollah war refined this as a needle

strike; soldiers are assets; wars are campaigns; deadly weapons are

systems, and so on.

In critical discourse analysis, the linguist Paul Chilton has

claimed that euphemism in discourse is the opposite of metaphor as

far as ideology is concerned: where metaphor replaces words and is

coercive and legitimizing, euphemism is suppressive and dissim-

ulating. He suggests that the linguistic strategies that effect euphe-

mism include not only the sort of lexical replacement listed abovebut also omission, passivization and nominalization.

See also: critical discourse analysis; metaphor

Further reading: Chilton 2003; Stockwell 2002a.

FINITE

Marked for tense. In many (not all) languages, a single verb may

exhibit a number of different forms serving different grammatical

functions, and these forms are often divided into two types called

finite and non-finite. English, with its small number of grammatical

markings, is not the ideal language to illustrate the difference, since

FINITE

89

some of the non-finite forms look just like some of the finite forms.

But let’s try.

A finite form is always marked for tense, and it also carries

agreement, insofar as English has any agreement. Consider the verbsmoke. Now this citation form (or lemma) of the verb, the form we

use to name it and to enter it in a dictionary, is a non-finite form,

the so-called infinitive. But the form smokes, as in Susie smokes, is

finite: that -s on the end tells us that this form is marked for present

tense and that it agrees with a third-person singular subject. More-

over, it is the only verb-form in the sentence, and only a finite form

can stand as the only verb in a sentence. By the same reasoning, the

form smoke which occurs in I smoke and they smoke is also finite,even though it carries no overt marking at all.

Also finite is the smoked of Susie smoked, which bears the past-

tense suffix -ed and is again the only verb in the sentence. But

present-tense smokes and smoke and past-tense smoked are the only

finite forms the verb smoke has.

A non-finite form is not marked for tense, it shows no agreement,

and it cannot be the only verb in a sentence. An example is the

form smoking, which has several functions. In Susie is smoking, itcombines with the auxiliary is, which itself is finite. In I have often

seen Susie smoking, it functions as a kind of complement (the finite

form here is have). In Smoking a guilty cigarette, Susie pondered the

ruins of her love life, it introduces a kind of modifier (the finite form

here is pondered). In Smoking is bad for you, it functions as a kind

of nominalization, the kind called a gerund (the finite form is is).

Also non-finite is smoked when it functions as a past participle, asin Susie has smoked since she was fifteen (the finite forms are has

and was), or as a passive participle, as in Cigars are rarely smoked by

women (the finite form is are). The infinitive smoke, also non-finite,

occurs in constructions like Susie can’t smoke in her office (the finite

form is can) and Susie wants to smoke (the finite form is wants).

If, as is sometimes done, we choose to regard sequences like is

smoking (as in Susie is smoking) and has smoked (as in Susie has

smoked since she was fifteen) as single verb-forms, then these are finite,

since the first element is finite. It is more usual, though, to analyse suchsequences into their finite and non-finite components. But not all such

sequences are finite: in Wanting to smoke a quick cigarette, Susie

made a rush for the balcony, the sequence wanting to smoke contains

no finite forms at all (the only finite form in the sentence is made).

A clause or a sentence containing a finite verb-form is itself finite.

Thus, Susie smokes is a finite sentence, while both clauses are finite in

FINITE

90

Susie always smokes a cigarette after she finishes dinner. But, inHaving

finished her dinner, Susie decided to smoke a cigarette, only Susie

decided to smoke a cigarette is finite, while having finished her dinner

and to smoke a cigarette contain no finite forms and are non-finite.

See also: tense

Further reading: Hurford 1994; Jackson 2002; Jeffries 2006.

FOCUS

Singling out some particular element of a sentence or an utterance

as representing the most important new information. Consider the

utterance Susie needs a holiday. Here it is not obvious that attention

is being drawn to any particular part of the utterance. But now add

stress: SUSIE needs a holiday. Now the utterance clearly means

‘The one who needs a holiday is Susie, and not somebody else’, and

we say that Susie is in focus. If we stress something else, then that

element is placed in focus: Susie needs a HOLIDAY (it’s not some-thing else that Susie needs).

In spoken English, we can always put a particular element in

focus by stressing it. But both spoken and written English have

another device for placing an element in focus: the use of any of

several types of cleft. So, we might say, or write, It’s Susie who

needs a holiday (placing Susie in focus), or What Susie needs is a

holiday (placing a holiday in focus). Also common, especially in

traditional music lyrics and in spoken discourse like sports com-mentaries, is topic headlining to create focus: My true love he is

handsome; Their defending is atrocious, the Londoners. However,

focus must be clearly distinguished from topic: even professional

linguists have been known to confuse these terms.

See also: given/new; topic

Further reading: Greenbaum and Quirk 1990.

FOLK LINGUISTICS

Speakers’ beliefs about their language or about language generally.

In any given speech community, speakers will usually exhibit many

FOLK LINGUISTICS

91

beliefs about language: that one language is older, more beautiful,

more expressive or more logical than another – or at least more

suitable for certain purposes – or that certain forms and usages are

‘correct’ while others are ‘wrong’, ‘ungrammatical’ or ‘illiterate’.They may even believe that their own language was a gift from a

god or a hero.

Such beliefs rarely bear any resemblance to objective reality,

except insofar as those beliefs create that reality: if enough English-

speakers believe that ain’t is unacceptable, then ain’t is unacceptable,

and, if enough Irish-speakers decide that English is a better or more

useful language than Irish, then they will speak English, and Irish

will die.It is because of facts like these that some, especially socio-

linguists, are now arguing that folk-linguistic beliefs should be

taken seriously in our investigations – in great contrast to the usual

position among linguists, which is that folk beliefs are no more

than quaint bits of ignorant nonsense.

See also: language myths; prescriptivism

Further reading: Crystal 1997.

FOREGROUNDING

A psycholinguistic term used in stylistics to refer to the sense that

certain elements in a discourse are more prominent than others.

Foregrounding often depends on stylistic deviance, which in turninvolves a sense of a default or background norm from which the

foregrounded element can emerge. Foregrounding is partly a function

of linguistic patterning and partly a matter of perception, of course.

In cognitive linguistics, the phenomenon of foregrounding is

described in terms of figure and ground relationships. Most pre-

positional meanings are reducible to image-schemas involving the

movement of a trajector against a landmark (figure and ground,

respectively). However, foregrounding is an important feature incognitive linguistics at the higher levels of text comprehension as

well.

See also: landmark; trajector

Further reading: Langacker 1987–91; Stockwell 2002b; van Peer 1986.

FOREGROUNDING

92

FORENSIC LINGUISTICS

A relatively recent discipline that applies linguistic techniques as a

means of establishing facts in criminal or detective cases. For exam-ple, criminals have been tracked down on the basis of their accent

signalling where they come from; semantic ambiguities have been

highlighted to introduce doubt into criminal convictions; patterns

in written texts have been used as part of the psychological profiling

of suspects. Forensic linguists explore the language of the criminal

court system, policing, and also matters of authorial attribution.

See also: applied linguistics

Further reading: Cotterill 2002; Gibbons 2003; McMenamin 2002; Shuy2005, 2006.

FORMAL GRAMMAR

A grammar which is expressed in terms of notational rules andtransformational algorithms. Generative grammars are the best-

known examples of formal grammar. The rules of the grammar

must be self-contained (that is, they should not reach outside their

own parameters to point to extraneous or social knowledge), the

notational system must consist of a finite set of elements, and the

algorithmic rules for transforming strings into other strings should be

finite and complete. Working in reverse, a computer program which

is based on these formal rules constitutes a parser of the language. Nocomputer program yet exists which is able to parse a language with

the absolute accuracy of a human being; this possibly points to a

fundamental theoretical flaw in the notion of formal grammars.

See also: generative grammar; grammar

Further reading: Antony and Hornstein 2003; Chomsky 1977, 2000.

FRAME

A term with a number of senses in different sub-disciplines of lin-

guistics. In various grammars, a frame refers to the syntactic fram-

ing of a class of words as a structural context: Suzie found – packet

FRAME

93

can frame a word from the class of determiners (the, a, her), for

example. In setting out the case system of a verb in a case language,

the list of possible cases for a verb is a case frame. In semantics and

cognitive linguistics, frames are sets of knowledge that are used toprovide interpretive information about a piece of language, some-

times called knowledge frames or schemas.

See also: case; cloze; schema

Further reading: Chafe 1970, 1994; Duchan 2003; Fillmore 1987.

FUNCTIONALISM

Any approach to the description of language structure which atta-

ches importance to the purposes to which language is put. Many

approaches to linguistics focus entirely on the purely structural

characteristics of languages, ignoring the possible functions of lan-

guage, and this approach has been very rewarding. But a large

number of linguists have preferred to combine the investigation ofstructure with the investigation of function; an approach which

does this is a functionalist approach.There are very many functionalist approaches which have been

put forward, and they are often very different from one another.

Two prominent ones are Role-and-Reference Grammar (RRG),

developed by William Foley and Robert Van Valin, and Systemic

Linguistics (SL), developed by Michael Halliday. RRG approaches

linguistic description by asking what communicative purposes needto be served and what grammatical devices are available to serve

them. SL is chiefly interested in examining the structure of a large

linguistic unit – a text or a discourse – and it attempts to integrate a

great deal of structural information with other information (social

information, for example) in the hope of constructing a coherent

account of what speakers are doing.

Functionalist approaches have proved fruitful, but they are

usually hard to formalize, and they often work with ‘patterns’,‘preferences’, ‘tendencies’ and ‘choices’, in place of the explicit rules

preferred by non-functionalist linguists.

See also: structuralism; Systemic Linguistics

Further reading: Bloor and Bloor 2004; Givon 1995; Halliday 2004;Siewierska 1991; Thompson 2004.

FUNCTIONALISM

94

FUNCTIONS OF LANGUAGE

The various purposes to which language may be put. We often tend

to assume that ‘the function of language is communication’, butthings are more complicated than that. Language serves a number

of diverse functions, only some of which can reasonably be regar-

ded as communicative. Here are some of the functions of language

which we can distinguish:

1 We pass on factual information to other people.

2 We try to persuade other people to do something.

3 We entertain ourselves or other people.

4 We express our membership in a particular group.

5 We express our individuality.

6 We express our moods and emotions.

7 We maintain good (or bad) relations with other people.8 We construct mental representations of the world.

All of these functions are important, and it is difficult to argue that

some of them are more important, or more primary, than others. For

example, studies of conversations in pubs and bars have revealedthat very little information is typically exchanged on these occasions

and that the social functions are much more prominent. Of course, a

university lecture or a newspaper story will typically be very different.

This diversity of function has complicated the investigation of the

origin and evolution of language. Many particular hypotheses about the

origin of language have tended to assume that just one of these diverse

functions was originally paramount, and that language came into

being specifically to serve that one function. Such assumptions arequestionable, and hence so are the hypotheses based upon them.

Proponents of functionalism are often interested in providing

classifications of the functions of languages or texts; see under

Systemic Linguistics for a well-known example.

See also: qualitative approach; Systemic Linguistics

Further reading: Bloor and Bloor 2004; Crystal 1997.

GAP

The absence in a sentence of a linguistic element which is in some

sense logically required. Though linguists had long been aware of

GAP

95

the existence of ellipsis, in the 1960s the American linguist Noam

Chomsky and his students began to pay particular attention to

certain types of ‘missing’ elements in the syntactic structures of

sentences. These missing elements were dubbed gaps. Here are a fewexamples of English sentences containing gaps; as is conventional,

the position of each gap is marked with the symbol e (for ‘empty’):

Susie wants e to buy a car (to buy has no subject); Susie is hard to

please e (please has no object); Who were you talking to e? (to has

no object); Susie bought a necklace and Zelda e a bracelet (the

second clause has no verb); Rod gave the museum a T-shirt and

Elton e e a pair of glasses (the second clause lacks both a verb and

the phrase the museum).The behaviour of gaps has increasingly been seen as crucial in

formulating adequate theories of grammar, and recent theories of

grammar often provide specific machinery for treating them.

Recently the name empty category has often been preferred to gap,

at least for those cases in which the ‘missing’ element is a noun

phrase.

See also: anaphor

Further reading: Barss 2003; Huddleston 1984.

GENDER

The classification of nouns into two or more classes with different

grammatical properties. In many of the world’s languages, all thenouns are divided into two or more classes which require different

grammatical forms on the noun and/or on certain other words

grammatically linked with the noun or nouns in particular sen-

tences. German, for example, has three gender classes (termed

masculine, feminine and neuter), which require different forms for

associated determiners and adjectives. Thus, ‘the table’ is der Tisch,

‘the pen’ is die Feder, and ‘the book’ is das Buch, where der, die, and

das are all different forms of ‘the’; ‘an old table’ is ein alter Tisch,‘an old pen’ is eine alte Feder, and ‘an old book’ is ein altes Buch.

It is important to realize that grammatical gender need have

nothing to do with sex. The word comes from the Latin genus,

meaning ‘kind’. In German (and other European languages), there

is a noticeable (but imperfect) correlation between sex and gender

assignment, but also some oddities (French feminine la barbe

GENDER

96

‘beard’); however, most nouns denote things that have no sex, and

yet they must still be assigned to a gender. In many other gender

languages, sex plays no part at all in gender assignment.

A gender language must have at least two gender classes, but itmay have more – eight, ten, or possibly even more. The native

Australian Dyirbal, for example, has four classes: masculine and

animate objects; feminine and dangerous objects; edible objects;

others. In some gender languages, we can often guess from the form

of a noun which gender it belongs to; in others, we can often guess

from its meaning which gender it belongs to; in very many lan-

guages, however, we cannot guess, because gender assignment is

arbitrary. In German, for example, a noun which denotes a male ora female usually (not always) goes into the der gender or the die

gender, respectively, and nouns with certain endings usually go into

a predictable gender. After that, though, the gender of the remain-

ing nouns is impossible to guess. In Navaho, nouns denoting

humans usually go into one gender, nouns denoting round things

into a second gender, nouns denoting long stiff things into a third

gender, and so on, but not all nouns can have their gender guessed

in this way.English, it is worth pointing out, has no gender. We have a few

sex-marked pronouns like he and she, and a few sex-marked nouns

like duke and duchess (and many that are becoming disused: baker/

baxter; actor/actress; jew/jewess), but we have no grammatical

gender.

Sociolinguists (and others) often use the term gender in a very

different way, to indicate the socially performed roles of masculi-

nities and femininities, with only an indirect relationship to biolo-gical sex. This usage must be carefully distinguished from the

strictly grammatical sense of the term.

See also: grammatical category

Further reading: Corbett 1991; Hurford 1994; Trask 1995; and on thesociolinguistic sense: Cameron and Kulick 2003; Coates 1998, 2004;Connell 1995; Holmes and Meyerhoff 2003; Mills 2003.

GENERATIVE GRAMMAR

A grammar of a particular language which is capable of defining all

and only the grammatical sentences of that language. The notion of

GENERATIVE GRAMMAR

97

generative grammar was introduced by the American linguist

Noam Chomsky in the 1950s, and it has been deeply influential.

Earlier approaches to grammatical description had focused on

drawing generalizations about the observed sentences of a language.Chomsky proposed to go further: once our generalizations are

accurate and complete, we can turn them into a set of rules which can

then be used to build up complete grammatical sentences from scratch.

A generative grammar is mechanical and mindless; once con-

structed, it requires no further human intervention. The rules of the

grammar, if properly constructed, automatically define the entire set

of the grammatical sentences of the language, without producing

any ungrammatical garbage. Since the number of possible sentencesin any human language is infinite, and since we do not want to

write an infinitely long set of rules, a successful generative grammar

must have the property of recursion: a single rule must be allowed

to apply over and over in the construction of a single sentence.

Chomsky himself defined several quite different types of gen-

erative grammar, and many other types have more recently been

defined by others. A key characteristic of any generative grammar is

its power: the larger the number of different kinds of grammaticalphenomena the grammar can handle successfully, the more power-

ful is the grammar. But – and this is a fundamental point – we do

not want our grammars to have limitless power. Instead, we want

our grammars to be just powerful enough to handle successfully the

things that actually happen in languages, but not powerful enough

to handle things that do not happen in languages.

Within certain limits, all the different kinds of generative gram-

mar can be arranged in a hierarchy, from least powerful to mostpowerful; this arrangement is called the Chomsky hierarchy. The

goal of Chomsky’s research programme, then, is to identify that

class of generative grammars which matches the observed proper-

ties of human languages most perfectly. If we can do that, then the

class of generative grammars we have identified must provide the

best possible model for the grammars of human languages.

Two of the most important classes of generative grammars so far

investigated are (context-free) phrase structure grammar and trans-formational grammar. The second is far more powerful than the

first – and arguably too powerful to serve as an adequate model for

human languages – while the first is now known to be just slightly

too weak (and has been modified).

(Special note: in recent years, Chomsky and his followers

have been applying the term generative grammar very loosely to the

GENERATIVE GRAMMAR

98

framework called Government-and-Binding Theory [GB], but it should

be borne in mind that GB is not strictly a generative grammar in

the original sense of the term, since it lacks the degree of rigorous

formal under-pinning which is normally considered essential in agenerative grammar.)

See also: phrase-structure grammar; transformational grammar

Further reading: Antony and Hornstein 2003; Bach 1974; Cook andNewson 1996; Lyons 1991.

GENETIC HYPOTHESIS OF LANGUAGE

The hypothesis that the human language faculty is rooted in our

genes. This hypothesis holds that our distinctive language faculty is

a trait which we have evolved over time, just like our upright pos-

ture and our opposable thumb. According to this view, language

just grows in children, much as their teeth grow, except that lan-

guage learning requires exposure to speech; that is, the hypothesissees our language faculty as a distinct and specific part of our

genetic endowment. It is seemingly supported by the nature of cer-

tain genetically based disabilities, which disrupt language while

affecting little else, or which leave language largely unaffected while

disrupting most other cognitive abilities. It is perhaps further sup-

ported by the existence of the astonishing language instinct in chil-

dren. While controversial, this hypothesis is now widely accepted by

linguists. The innateness hypothesis is a more specific version of it.Nevertheless, the genetic hypothesis has been vigorously criticized

by the British linguist-turned-computer-scientist Geoffrey Sampson

and by the American psychologist Elizabeth Bates and her colleagues.

See also: autonomy; innateness hypothesis; language instinct

Further reading: Bates 1976; Bates et al. 1979; Bates et al. 1988; Crowley1997; Macwhinney and Bates 1989; Pinker 1994; Sampson 1997.

GENETIC RELATIONSHIP

The relationship between languages which share a common ancestor.

Living languages are always changing, and, when a single language

GENETIC RELATIONSHIP

99

extends over a significant geographical area, different changes inevi-

tably occur in different places. Over time, then, the original lan-

guage breaks up, first into a continuum of regional dialects, then

eventually into several quite distinct languages. The several lan-guages which result are the daughter languages of their common

ancestor, and these daughters are genetically related.Over millennia, this splitting may be repeated again and again,

and that single ancestral language may thus give rise to a sizable

language family. All the languages in the family are genetically

related; those which share a more recent common ancestor are

more closely related than those whose last common ancestor is

more distant.One of the principal goals of historical linguistics is the identifi-

cation of genetic relationships. This is often easy when the lan-

guages in question are closely related – that is, when their last

common ancestor was spoken not more than two or three millennia

ago. More distant genetic links, resulting from more remote

common ancestry, are more difficult to identify with certainty and

require careful analytical procedures to avoid being misled by

chance resemblances and ancient instances of language contact.Eventually, at some time-depth, genetic links become impossible to

identify at all, because the ceaseless processes of linguistic change

will obliterate all traces of a common origin, or at least render them

unrecognizable amid the background noise. Features of language

change such as borrowing, language planning and prescriptivism also

serve to disrupt the neat family-tree metaphor involved here.

It is possible that all languages are ultimately descended from a

single common ancestor – the ancestral speech of the first humans –and hence that all are genetically related, but we will never know

about it.

See also: historical linguistics; language family; proto-language; recon-struction

Further reading: Lehmann 1992; McMahon 1994; Trask 1994, 1996.

GENRE

A historically stable variety of text with conspicuous distinguishing

features. The concept of genre is shared by (at least) linguistics,

anthropology and literary criticism. Its study is well established but

GENRE

100

contentious, and figures ranging from the Russian linguist Roman

Jakobson to the Russian Marxist literary critic Mikhail Bakhtin

have made important contributions.

The key fact about a given genre is that it has some readilyidentifiable distinguishing features that set it off markedly from

other genres, and that those features remain stable over a sub-

stantial period of time. In most cases, a particular genre also occu-

pies a well-defined place in the culture of the people who make use

of the genre.

Among the genres familiar to most of us are lyric poetry, religious

liturgy, legal documents, proverbs, fairy tales, scholarly mono-

graphs, and news stories. Other societies may present further types,such as the illness-curing chants of Mayan shamans and the oral

epic poems of Serbian or ancient Greek bards. Very often mastery

of a particular genre is seen as a requirement for a certain profes-

sion; this is so for lawyers, bards, academics, shamans, scientists

and physicians, among others.

It is characteristic of every genre that the outward form of

expression is of vital significance, and at least as important as the

content; in some cases the form may actually be more importantthan the content, as is true of many types of poetry, such as French

villanelles and Japanese haiku. In many communities, song and

verse genres are characterized by such features as the use of totally

different words from the everyday ones and the requirement that no

word may ever be repeated. But even a scientific paper is subject to

rigid rules of form: the order of presentation must be background/

procedure/results/interpretation/conclusions; the paper must be

written in an impersonal third person; and all mistakes, accidentsand dead ends that cropped up during the work must be silently

omitted. A chemist who volunteered in a paper ‘At this point I

dropped the beaker on the floor and had to start over’ would not

get his or her paper published.

See also: text

Further reading: Bakhtin 1968, 1984; Bhatia 1993; Foley 1997.

GIVEN/NEW

A way of classifying the elements of a sentence according to their

information content. Most utterances are not produced in isolation:

GIVEN/NEW

101

instead, each is produced in some context involving what has been

said previously and what is known to, or believed by, the speaker

and the listener. As a result, it is often the case that some part of an

utterance serves only to tie it to this context, while another partintroduces some kind of new information. We therefore speak of

the given/new distinction.

The given part of an utterance represents the part which is

already familiar to the listener in one way or another, while the newpart represents the main contribution of the utterance. Consider the

following exchange. Mike: I don’t know the woman in the white dress.

Susie: Oh, she’s the new Professor of Psychology. Here Susie’s reply

can be analysed into the given part she and the new part is the new

Professor of Psychology.

The analysis of sentences and utterances in terms of their orga-

nization of information was pioneered by the linguists of the

Prague School in the early twentieth century, especially by the

Czech Vilem Mathesius, under the name functional sentence per-spective. Instead of given and new, Mathesius used the terms themeand rheme, and these are still in use today, especially by the pro-

ponents of Systemic Linguistics, though the terms are used here in aslightly specialized way. Still other linguists prefer the terms topic

and comment in the same senses.

See also: topic

Further reading: Brown and Miller 1991; Greenbaum and Quirk 1990;Thompson 2004.

GOVERNMENT

The grammatical phenomenon in which the presence of a particular

word in a sentence requires a second word which is grammatically

linked with it to appear in a particular form. Most English personal

pronouns occur in two different case forms, the nominative and the

objective. Examples: I/me, she/her, they/them. When a prepositiontakes one of these pronouns as its object, that pronoun must appear

in its objective form: with me, not *with I; for her, not *for she. We

say that the preposition governs the case of its object, or simply that

the preposition governs its object. What this means is that it is the

very presence of the preposition which requires the objective case.

This is not agreement: it makes no sense to say that the pronoun is

GOVERNMENT

102

‘agreeing’ with the form of the preposition, because an English

preposition has only a single form.

Government can be more complex. In German, for example,

there are several different cases, and some prepositions govern onecase, others another. For example, mit ‘with’ governs the dative case,

and so German requires mit mir ‘with me’, with the dative case-

form mir of the pronoun ich ‘I’. But the preposition fur ‘for’ gov-

erns the accusative case, and so ‘for me’ is fur mich, with accusative

mich. Each preposition in German governs some particular case,

and a learner simply has to learn which prepositions require which

cases.

Verbs can also govern case-forms. In Basque, for example, aparticular verb may govern objects in any of several cases. Most

govern the absolutive case; for example, ikusi ‘see’ does so: neska

ikusi dut ‘I saw the girl’ (neska ‘the girl’, absolutive). But some

govern the dative, such as lagundu ‘help’: neskari lagundu diot ‘I

helped the girl’ (neskari, dative). And a few govern the instru-

mental, such as gogoratu ‘remember’: neskaz gogoratu naiz ‘I

remembered the girl’ (neskaz, instrumental). In English, all verbs

govern objects in the objective case.For the somewhat special case of gender government, see under

agreement. In the Government-and-Binding Theory, the concept of

government is generalized and extended in certain ways that are

central to the machinery of that framework.

See also: agreement; dependency; subcategorization

Further reading: Gleason 1961.

GOVERNMENT-AND-BINDING THEORY

A particular theory of grammar, the descendant of transformational

grammar. During the 1960s and 1970s, Noam Chomsky’s transfor-

mational grammar went through a number of substantial revisions.

In 1980, Chomsky gave a series of lectures in Pisa outlining a dra-matic revision of his ideas; these lectures were published in 1981 as

a book, Lectures on Government and Binding. The new framework

presented there became known as the Government-and-BindingTheory (GB) or as the Principles-and-Parameters approach.

GB represents a great departure from its transformational

ancestors; while it still retains a single transformational rule, the


103

framework is so different from what preceded it that the name

‘transformational grammar’ is not normally applied to it.

As the alternative name suggests, GB is based squarely upon two

ideas. First, the grammars of all languages are embedded in a uni-versal grammar, conceived as a set of universal principles applying

equally to the grammar of every language. Second, within universal

grammar, the grammars of particular languages may differ only in

small and specified respects; these possible variations are conceived

as parameters, and the idea is that the grammar of any single lan-

guage will be characterized by the use of a particular setting for

each one of these parameters. The number of available settings for

each parameter is small, usually only two or three.GB is a modular framework. Its machinery is divided up into

about eight distinct modules, or components. Each of these modules

is responsible for treating different aspects of sentence structure,

and each is subject to its own particular principles and constraints.

A sentence structure is well formed only if it simultaneously meets

the independent requirements of every one of the modules. Two of

those modules – those treating government and binding (the possi-

bility that two noun phrases in a sentence refer to the same entity) –give GB its name.

Just like transformational grammar, GB sees every sentence as

having both an abstract underlying structure (the former deep

structure, now renamed D-structure) and a superficial structure (the

former surface structure, now renamed S-structure). There is also a

third level of representation, called logical form (LF). Certain

requirements apply to each one of these three levels, while further

requirements apply to the way in which the three of them are related.The motivation for all this, of course, is the hope of reducing the

grammars of all languages to nothing more than minor variations

upon a single theme, the unvarying principles of universal gram-

mar. But the task is far from easy, and Chomsky, confronted by

recalcitrant data, has been forced into the position of claiming that

the grammar of every language consists of two quite different parts:

a core – which alone is subject to the principles of universal

grammar – and a periphery – consisting of miscellaneous language-specific statements not subject to universal principles. This ploy has

been seen by critics as a potentially catastrophic retreat from the

whole basis of the Chomskyan research programme.

GB was an abstract framework to begin with, but it has become

steadily more abstract, as its proponents, confronted by trouble-

some data, have tended to posit ever greater layers of abstraction,


104

in the hope of getting their universal principles to apply successfully

at some level of representation. Critics have not been slow to see

this retreat into abstraction as a retreat from the data altogether,

that is as an attempt to shoehorn the data into a priori principleswhich themselves are sacrosanct. The more outspoken critics have

declared the GB framework to be more a religious movement than an

empirical science. Nevertheless, GB has for years been by far the most

influential and widely practised theory of grammar in existence.

Recently, however, Chomsky has, to general surprise, initiated the

Minimalist Program (original US spelling), in which almost all of

the elaborate machinery of GB is rejected in favour of a very dif-

ferent approach.

See also: transformational grammar

Further reading: Antony and Hornstein 2003; Cook and Newson 1996;Cowper 1992; Culicover 1997; Haegeman 1994; Horrocks 1987; Ouhalla1994; Sells 1985.

GRAMMAR

The rules for constructing words and sentences in a particular lan-

guage, or the branch of linguistics studying this. Every language has

a grammar. Where this is set out explicitly in published linguistic

descriptions, in textbooks, and taught through the school system, it

can be said to be codified and standardized. Of course, those many

languages which have never been subject to professional linguists’detailed scrutiny also operate according to their grammars. The

linguistic study of grammar is conventionally divided into two

parts: morphology – the study of word structure – and syntax – the

study of sentence (or utterance) structure.

The tradition of studying grammar is venerable: the ancient

Indians, the ancient Greeks and Romans, and the medieval Chi-

nese, Arabs and Jews all did important grammatical work on their

favourite languages, and the Port-Royal grammarians in seventeenth-century France were already contemplating grammar from a uni-

versalist point of view. But the rise of modern linguistics in the

early twentieth century gave new impetus to the study of grammar;

by the 1930s and 1940s the American Leonard Bloomfield and his

successors were doing important work in morphology, and in the

1950s Noam Chomsky made the study of syntax one of the most

GRAMMAR

105

prominent of all areas of linguistics, by introducing the new approach

called generative grammar and by reviving the search for universal

grammar.

Approaches to the study of grammar are many and various.Pre-twentieth-century approaches represent traditional grammar,

while most twentieth-century approaches are varieties of structur-

alism. The more formal approaches developed since the 1950s are

known as theories of grammar; among the more prominent ones

are the several versions of phrase-structure grammar, Lexical-

Functional Grammar, and transformational grammar with its des-

cendant Government-and-Binding Theory. Among the approaches

embedded within functionalism, the most prominent is SystemicLinguistics.

See also: generative grammar; morphology; syntax; universal grammar

Further reading: Brown and Miller 1991; Hurford 1994; Kroeger 2005;Pinker 1994.

GRAMMATICAL CATEGORY

A linguistic category which has the effect of modifying the forms of

some class of words in a language. The words of every language are

divided up into several word classes, or parts of speech, such as

nouns, verbs and adjectives. It often happens that the words in a

given class exhibit two or more forms used in somewhat different

grammatical circumstances. In each such case, this variation inform is required by the presence in the language of one or more

grammatical categories applying to that class of words.

English nouns are affected by only one grammatical category, that

of number: we have singular dog but plural dogs, and so on for most

(but not all) of the nouns in the language. These forms are not

interchangeable, and each must be used always and only in specified

grammatical circumstances. And here is a key point: we must

always use a noun in either its singular form or its plural form, evenwhen the choice seems irrelevant; there is no possibility of avoiding

the choice, and there is no third form which is not marked one way

or the other. This is typically the case with grammatical categories.

English pronouns sometimes vary for case, as with I/me, she/her

and they/them, and again only one of the two forms is generally

possible in a given position. English nouns lack the category of

GRAMMATICAL CATEGORY

106

case, but nouns in Latin, German, Russian and many other lan-

guages do vary in form for case.

English adjectives vary for the grammatical category of degree, aswith big/bigger/biggest, and yet again only one of the three forms ispossible in a given position.

English verbs exhibit the category of tense, as with love/loved,

work/worked, see/saw, take/took, sit/sat and drive/drove. (Quite a few

of these are irregular, but that does not matter: what matters is that

the first form is required in certain circumstances but the second in

others, and so we have I see her [now] but I saw her [in the past].)

(English verbs have other forms, of course, but these other forms

are either not marked for tense, or not just for tense; instead theyare marked for other grammatical categories like aspect and voice.)

Some grammatical categories, like number and tense, are extre-

mely widespread in the world’s languages (though by no means

universal), while others are unusual and confined to a few lan-

guages. For example, some North American languages have a

grammatical category of visibility, by which nouns and pronouns

must be explicitly marked to indicate whether or not the speaker

can see the things they refer to at the moment of speaking. Andmany languages have the category of evidentiality, by which every

statement must be overtly marked to show the source of the

speaker’s information: ‘I saw it myself’, ‘Somebody told me’, ‘I have

inferred this from evidence’, and sometimes further distinctions.

See also: aspect; case; deictic category; gender; mood; number; person;tense; voice

Further reading: Crystal 1997; Gleason 1961; Lyons 1968; Palmer 1971.

GRAMMATICAL RELATION

Any one of the ways in which a noun phrase may be related to a

verb in a particular sentence. Grammatical relations have been

recognized as fundamental since ancient times. Though they werenoticeably ignored during the early days of generative grammar,

they have once again come to be viewed as an essential part of the

grammatical structure of a sentence.

Grammatical relations, which are sometimes called grammaticalfunctions, are also surprisingly difficult to define explicitly. The most

familiar grammatical relation is that of subject. In English, the

GRAMMATICAL RELATION

107

subject of the sentence usually comes first, or at least before the verb,

and it is the only thing the verb ever agrees with (English doesn’t

have much agreement, of course). The subjects of the following

sentences are bracketed: [Susie] smokes; Carefully [she] poured the

wine; [My girlfriend’s parents] are visiting us; [Most of my students]

drink; [That you are worried] is obvious. The part of the sentence

that follows the subject is the predicate (in one sense of that term).

A noun phrase that follows a verb is in most cases a direct object.

Examples: She likes [me]; Susie wants [a new car]; Susie has visited

[most of the countries in Europe]. But, if the verb is one of a small

group including be and become, the following noun phrase is not a

direct object but a predicate nominal. Examples: Susie is [the cle-

verest person I know]; Susie became [an atheist].

An oblique object is less directly connected to the verb. In English,

an oblique object surfaces as the object of a preposition, though

some other languages use case-endings for the same purpose.

Examples: I went to Spain with [Lisa]; The cat is under [the bed].

Another traditional grammatical relation is the indirect object,

but it is not clear whether or not indirect objects exist in English.

Traditional grammarians would say that Lisa is an indirect objectin both Mike gave this book to Lisa and Mike gave Lisa this book,

and some linguists agree. However, in the first it is hard to see that

Lisa is anything other than an ordinary oblique object. In the

second, somewhat surprisingly, Lisa is arguably a direct object: note

the corresponding passive Lisa was given this book by Mike.

See also: grammar; parsing; syntax

Further reading: Brown and Miller 1991; Hurford 1994.

GRAPHEME

A single character in a recognized writing system. Every established

writing system necessarily makes use of some set of written char-

acters. Depending both upon the nature of the system used and onthe facts of the language being written, this number may range

from a mere handful up to many thousands.

At the simplest level of analysis, for example, the version of the

roman alphabet used for writing English makes use of some eighty-

odd graphemes: the twenty-six capital letters <A>, <B>, <C>, . . . ,the twenty-six small letters <a>, <b>, <c>, . . . , the ten digits <0>,

GRAPHEME

108

<1>, <2>, . . . , an assortment of punctuation marks like <.>, <,>,

<?> and <;>, and the blank space < >. (It is conventional to

enclose a grapheme in angle brackets.)

A more sophisticated analysis of English writing might prefer toset up some additional graphemes, notably the digraphs used for

writing single sounds, such as the <sh> of ship, the <ch> of chip,

the <th> of both thin and then, the <ng> of sing and the <ea> and

<ee> of bread and reed. Indeed, advocates of spelling reform (such

as, most famously, Noah Webster, Mark Twain and George Ber-

nard Shaw) have recommended single joined graphemes to repre-

sent these phonemes.

Some other languages using the roman alphabet have additionalgraphemes involving diacritics, such as <c>, <s>, <n>, <e>, <a>

and <ø> (plus their capital versions); these are sometimes counted

as distinct letters of the alphabet and sometimes not, but they are

still distinct graphemes. A few languages even add further letters,

like Icelandic and Old English <æ> and German <ß>. The Arabic

alphabet has no capital letters, but most letters have two or even

three different graphic forms, depending on where they occur in a

word, and each of these different forms is a grapheme.The non-alphabetic Chinese writing system uses several thousand

graphemes for everyday purposes and thousands more for specialist

purposes, and the complex mixed system represented by the Egyp-

tian hieroglyphs used a total of nearly 5,000 graphemes.

The standard ASCII set of characters found on most computer

keyboards contains ninety-five graphemes, including such symbols

as <$>, <*>, <+>, <&> and <@>, which are not usually counted

as graphemes in the English writing system. (Incidentally, in lip-reading as used by some deaf people, sounds which look the same

in terms of mouth and face shape are called visemes: examples

include the group /p b m/ and /1 t d n/. Words like pet, met, bet,

bed, bell, Ben, men and pen cause difficulty for lip-readers.)

See also: orthography; writing system

Further reading: Coulmas 1996.

HEAD

That element in a phrase which is chiefly responsible for the nature

of that phrase. Every phrase in every language is built up from

HEAD

109

smaller units according to certain rather rigid rules. The several

different types of phrase are distinguished from one another to

some extent by differences in structure but mainly by the nature of

the item (usually a word) around which it is constructed; that itemis the lexical head, the headword, or simply the head, of the phrase,

and it usually provides the name for the kind of phrase built up

around it.

For example, the noun phrase the little girl in the blue dress is built

up around the noun girl as its head; the whole phrase denotes some

kind of girl. Likewise, the verb phrase sang quietly to herself is built

up around the verb sang as its head; the whole phrase denotes some

specific kind of singing. The adjective phrase pretty as a picture isheaded by the adjective pretty; the adverb phrase very slowly is

headed by the adverb slowly; and the prepositional phrase under the

bed is headed by the preposition under.

The concept of heads is an ancient one, but it largely dis-

appeared from linguistics during the earlier part of the twentieth

century. Since the 1970s, however, heads have once again come to

be seen as grammatically central, and many contemporary theories

of grammar assign them a very prominent place; one or two the-ories even take headedness to be the most important grammatical

notion of all.

See also: constituent structure; modifier; phrase; phrase-structuregrammar

Further reading: Brown and Miller 1991; Hurford 1994.

HISTORICAL LINGUISTICS

The study of language change and of its consequences. Historical

linguistics was the first branch of linguistics to be placed on a firm

scholarly footing. It is traditional to date the founding of the dis-

cipline to 1786, when the British amateur linguist Sir William Jones

famously pointed out the clear common ancestry of Greek, Latinand Sanskrit and hence of the existence of the vast Indo-European

family of languages, all of which descend from a single common

ancestor. At about the same time, however, several Hungarian lin-

guists were establishing that Hungarian must likewise share a common

origin with Finnish and several other languages, in a different

family now called Uralic.

HISTORICAL LINGUISTICS

110

Historical linguistics was vigorously developed throughout the

nineteenth century, chiefly by linguists who were German or trained

in Germany. Most of the attention was on comparative linguistics:the business of deciding which languages shared a common ances-try and hence which language families existed, of performing

reconstruction to work out the properties of unrecorded ancestral

languages (proto-languages), and of identifying the various changes

which had led each ancestral language to break up into its several

divergent daughters.

In the latter part of the nineteenth century, a number of younger

linguists decided that they had enough evidence to declare that

sound change was invariably regular – that is, that a given sound ina given context in a given language always changed in the same

way, without exception. This Neogrammarian hypothesis became the

orthodoxy in the field for the next hundred years, and it proved

very fruitful.

In the twentieth century, and especially in recent years, there has

been an explosion of interest in all aspects of language change. In

particular, linguists have been searching eagerly for principles gov-

erning language change: what makes some changes more likely thanothers? It has proved possible to study changes which are in pro-

gress in contemporary languages, including English, and such stu-

dies have turned up a number of startling phenomena, many of

which are clearly incompatible with the Neogrammarian hypoth-

esis. A key point has been the discovery of the crucial link between

variation and change. Historical linguistics has once again become

one of the liveliest areas in all of linguistics.

See also: comparative reconstruction; internal reconstruction; languagechange; language family; Saussurean paradox; sociolinguistics; systematiccorrespondence

Further reading: Crowley 1997; Fennell 2000; Hock and Joseph 1996;McMahon 1994; Trask 1996.

ICONICITY

A direct correlation between the form of a word and its meaning.

The overwhelming norm in languages is arbitrariness, by which the

form of a word bears no relation to its meaning. But there are certain

exceptions, and these exceptions exhibit varying degrees of iconicity.

ICONICITY

111

There are several types of iconicity. The most obvious is pho-naesthetic analogy, which includes onomatopoeia. An onomatopoeic

word is one which denotes a sound and which has a linguistic form

specifically designed to mimic that sound with some degree ofrecognizability. English examples include clink, meow, hiss, bang,

boom, hum, quack and woof. But even onomatopoeic words exhibit

a good deal of arbitrariness: the sound of a gunshot is represented

as bang in English, but as pum in Spanish, peng in German, and

dzast in Basque. Non-onomatopoeic phonaesthesia can be dis-

cerned perhaps in the Basque word tximeleta (roughly, chee-may-

LAY-tah) ‘butterfly’, or the English word ‘jittery’, though these

become debatable.Other forms of iconicity include the sequential order principle, by

which the order of events or the order of perceptual focus is reflec-

ted in the linguistic sequence; the quantity principle, which suggests

that linguistic complexity is matched to conceptual complexity; and

the proximity principle, which suggests that conceptual or emotional

distance is mirrored in linguistic distance. This last form of iconi-

city includes ‘distance’ measured not just in terms of the length of

the utterance but also in terms of distancing caused by modaliza-tion, hypothesis, negation, fictional narrativizing, and so on. The

study of how experience and language are related iconically is part

of cognitive linguistics.

See also: arbitrariness; cognitive linguistics; sound symbolism

Further reading: Anttila 1988; Hiraga 2005.

IDENTITY

Identity in relation to linguistics concerns the role of language in

providing a speaker with individuality and group membership. Every

time you open your mouth, you give other people important clues

about what sort of person you are: where you come from, what social

class you belong to, even your sex and age (for example, on thetelephone). This information says something both about your indi-

viduality and about the social, national and ethnic groups to which

you consider yourself to belong. For several decades now, socio-

linguists have realized that providing each speaker with an indivi-

dual and group identity is one of the most important functions of

language.

IDENTITY

112

An appreciation of this identifying function of language is crucial

in understanding many types of social and linguistic behaviour. One

of the most obvious of these is the seeming paradox that many

people consistently describe their own speech as ‘bad’ or ‘wrong’ or‘inferior’, and yet make no effort to change it towards the sort of

speech they explicitly describe as ‘better’. In fact, people have been

observed asserting one view of their own variety’s stigma, while

simultaneously encouraging and correcting their children’s speech

towards that stigmatized vernacular. In these cases, the maintenance

of a valued social identity overrides any other consideration.

The link between language and identity can involve entirely dif-

ferent languages. Welsh-speakers in Wales or Basque-speakers andCatalan-speakers in Spain may (and often do) regard their dis-

tinctive language as a central part of their identity and may deeply

resent pressures to abandon their ancestral language in favour of

the more prestigious English or Spanish. Rather than consenting to

becoming anonymous if slightly quaint speakers on the fringes of

the English-speaking or Spanish-speaking world, they prefer to see

themselves as part of a distinct people, with their own nation, their

own history, their own traditions, their own values and their owngoals; the most obvious outward sign of this distinct identity is

their language. The all-too-common failure to recognize or esteem

this identifying effect of language has led countless times to grief

and to major social, educational and political problems.

See also: social stratification of language

Further reading: Crystal 1997; de Fina 2003; Fishman 1999; Joseph2004.

IDEOLOGY

In linguistics, ideology refers to the set of beliefs underlying an

utterance or discourse. It is an especially important notion in

Systemic Linguistics and in critical discourse analysis. In everydaylanguage, ‘ideology’ and ‘ideological’ tend to have negative con-

notations, but these linguists regard every example of language in

use as having an ideological dimension. For example, an utterance

that describes an event in the world has to choose one of the many

possible lexicogrammatical ways in which that event can be enco-

ded: active, passive, focusing on one topic rather than another,

IDEOLOGY

113

foregrounding one perspective rather than another, choosing parti-

cular naming and address patterns rather than others, selecting a

level for formality, register, politeness, and so on. All of these

encode a range of ideological choices, and there is no neutralchoice. Choices which are seen as natural and neutral are usually

simply ideological realizations which the receiver agrees with.

See also: identity; minority language; national language; official language;power

Further reading: Bourne and Cameron 1989; Crowley 1996; Fairclough1995; Simpson 1993.

IDIOM

An expression whose meaning cannot be worked out from the

meanings of its constituent words. Even if you know the meanings

of all the words in the phrase let the cat out of the bag, you cannot

guess the idiomatic meaning of the whole expression: this you mustlearn separately. (It means, of course, to reveal something publicly

which is supposed to be a secret.) Such an expression is an idiom,

and English, like other languages, has lots of idioms. Among these

are buy a pig in a poke (commit oneself to an irrevocable course of

action without knowing the relevant facts), the tip of the iceberg

(the small visible part of a large problem), kick the bucket (die),

three sheets to the wind (drunk) and stick to one’s guns (refuse to

change one’s mind or give up).The meanings of all such idioms are unpredictable and must be

learned separately. Many such idioms are so familiar that native

speakers hardly realize they are using an idiom at all. Exposure to a

foreign language quickly reveals the true position: for example, the

Basque idiom Ez kendu babak altzotik is literally ‘Don’t take the

beans out of your lap’, but no non-speaker of Basque is likely to

guess the idiomatic meaning. (It means ‘Don’t get up on your high

horse’, ‘Don’t lecture me’, and perhaps you can now see the moti-vation for the idiom.)

Idioms usually derive from the social history of the language

community and thus offer a good guide to the cultural concerns of

that society. British English has many idioms from its seafaring past

(copper-bottom investment, money for old rope, cash on the nail, set

off on a different tack, got him over a barrel, between the devil and

IDIOM

114

the deep-blue sea, and many others including three sheets to the

wind). Idioms often, though not always, involve metaphor.

A linguistically fascinating fact about idioms is that some of

them (though not all of them) can undergo the ordinary syntacticprocesses of the language. For example, let the cat out of the bag

can appear in sentences like The cat has been well and truly let out

of the bag, or Several cats were let out of that particularly secret bag,

in which the idiom has been broken up and its parts scattered about

the sentence, and yet the idiomatic sense is still present. Such

findings pose interesting problems both of syntax and of

psycholinguistics.

See also: metaphor; vernacular

Further reading: Fernando 1996; Huddleston 1984; Langlotz 2006.

INDICATOR

In sociolinguistics, an indicator is a linguistic variable which doesnot carry any social significance for the speaker, unless their atten-

tion is explicitly drawn to it. Such features are typically indicative

of some social factor, though they are below the conscious aware-

ness of speakers. As such, they are very useful for sociolinguists.

Examples might include the systematic use of ‘dark /l/’ (a velarized

/l/) for a ‘light /l/’ (alveolar or dentalized) in many Scottish accents,

which most speakers can hear but are not normally explicitly aware

of; or the use of the non-standard ‘I’ll be there while midnight’ (=‘until midnight’) which many north-western English speakers use

without being conscious, until they travel elsewhere, that it is non-

standard.

See also: marker; stereotype

Further reading: Chambers 2003; Llamas et al. 2007; Stockwell 2002a.

INDIRECT OBJECT

In a sentence I gave him the ball, there are three arguments: the

subject (I), the direct object (the ball) and an indirect object (him)

which receives the action of the verb and is additional to the

INDIRECT OBJECT

115

obligatory direct object. In case languages, the indirect object is

usually marked in the dative case. In languages like English, it is

indicated in the word-order, or sometimes with the particle ‘to’. So

the following are allowed: I gave the ball to him, To him I gave the

ball. The following sound odd: *I him gave the ball, *I gave the ball

him, though the latter is possible in some dialects. There are dis-

agreements in different grammars over what counts as an indirect

object, and in generative grammar, only direct objects are

recognized – all other noun phrases are treated separately.

See also: argument; case; direct object

Further reading: Blake 2001; Jackson 2002; Jeffries 2006.

INDO-EUROPEAN

A vast language family. People have long been aware that certain

languages are strikingly similar to certain other languages. In the

late eighteenth century a few linguists began to realize that certainancient languages of Europe and Asia, notably Latin, Greek and

Sanskrit (in India), were so remarkably similar in their grammars

that they must share a common origin. This observation, famously

made by Sir William Jones in 1786, marks the official beginning of

the recognition of the Indo-European (IE) family. It was quickly

realized that Gothic (and the other Germanic languages), Old Per-

sian (and the other Iranian languages) and the Celtic languages also

shared the same common origin, as well as the Baltic and Slaviclanguages, and Albanian and Armenian. Over a century later, texts

written in several long-extinct languages were unearthed in Anato-

lia and central Asia; when deciphered, these too proved to be writ-

ten in ancient Indo-European languages: Hittite (and several other

Anatolian languages) in the first case and the Tocharian languages

in the second. A very few other ancient Indo-European languages

have turned up in inscriptions but are so poorly documented we

know little about them.By applying comparative reconstruction (which was largely devel-

oped and refined by its application here), linguists were eventually

successful in carrying out a substantial reconstruction of the single

language which was the remote ancestor of all these languages. This

ancestor is called Proto-Indo-European, or PIE. The speakers of

PIE were illiterate and left no records, but we nevertheless know a

INDO-EUROPEAN

116

great deal about the phonology, grammar and vocabulary of PIE.

For example, we are confident that a PIE root of the approximate

form *kwel- ‘revolve, turn’ (the asterisk indicates a reconstructed

form) is the source of Latin collum ‘neck’, Greek polos ‘pole’ –borrowed as English (north) pole – and Old Church Slavonic kolo

‘wheel’ (among others), and that a reduplicated form of this

*kwekwlo- is the source of Greek kuklos ‘circle, wheel’ – borrowed

as English cycle – Sanskrit cakra- ‘wheel’ and English wheel (Old

English hweowol).

We think PIE was probably spoken around 6,000 years ago, but the

place is unknown. Most people favour the south Russian steppes,

but others have argued for the Balkans, Anatolia, the Middle East,the Caucasus, central Asia and other locations. In any case, the IE

languages eventually spread over a huge area of Asia and most of

Europe, repeatedly breaking up into clusters of daughter languages

as they did so. In the process, the IE languages obliterated an

unknown number of earlier languages, including all the earlier lan-

guages of Europe except for Basque in the west and Finnish and its

relatives in the north. Until the eve of modern times, the Indo-

European family extended from India to western Europe (hence itsname), but the European colonial expansion has introduced IE

languages like English, Spanish, French, Portuguese and Russian

into vast areas of the globe, displacing many more languages in the

process. Today about half the world’s people speak IE languages,

even though the number of living IE languages is below 200.

See also: comparative reconstruction; language family; reconstruction

Further reading: Baldi 1983; Beekes 1995; Lehmann 1967, 1993; Lock-wood 1969, 1972; Mallory 1989; Szemerenyi 1996; Trask 1996.

INFINITIVE

Traditional grammar has the unmarked form of a verb expressed as

to go, to be, to do and calls these infinitives. However, no moderngrammar follows this practice. The infinitive is the non-finite form

of a verb, unmarked for tense, person, number or mood. Some

languages have a separate form for the infinitive (Spanish completar

‘to complete’, French venir ‘to come’); English uses an uninflected

stem (complete, come, be, do, go). Infinitives are most often used as

the complements of other verbs: He wanted to go, I was going to be

INFINITIVE

117

there by ten, To jump in the river would not be allowed. In the sen-

tence, He was going to go boldly where no one had gone before, any

syntactic test would show that go boldly where no one had gone

before is a constituent verb phrase, and there is no reason other thanhistorical ignorance to outlaw the ‘splitting of the infinitive’ and say

correctly He was going to boldly go where no one had gone before.

See also: finite; verb

Further reading: Jackson 2002; Hurford 1994.

INFLECTION

Variation in the form of a single word for grammatical purposes. In

many (not all) languages, a single word can assume any of several

different forms, or even dozens of different forms, the choice

depending on the grammatical context in which it is used. This is

inflection. A word may be inflected by adding affixes or by various

types of internal change.English has very little inflection, but it does have some. A typical

noun has only two grammatical forms: singular and plural (dog/

dogs, child/children, foot/feet). A typical verb has slightly more forms

than a noun; for example, write has write, writes, wrote, written,

writing, while love has only love, loves, loved, loving, and put has just

put, puts, putting. (The number is larger if we count syntactic forms

like has written as inflections, which we sometimes do but usually

don’t.) A typical adjective has three inflected forms: positive big,comparative bigger, superlative biggest.

Inflection is not universal. Vietnamese, for example, has no

inflection at all, and every word is completely invariable in form.

On the other hand, some North American languages have aston-

ishingly complex inflectional systems, in which a single verb may

appear in hundreds of different forms.

The key point about inflection is that applying it never gives you

a new word, but only a different form of the same word (thelemma). But derivation, in contrast, does produce new words which

have to be entered separately in a dictionary.

See also: derivation; grammatical category; morphology

Further reading: Bauer 1988; Brown and Miller 1991.

INFLECTION

118

INNATENESS HYPOTHESIS

The hypothesis that children are born knowing what human lan-

guages are like. It is obvious that particular languages are notinnate and must be learned. Any child, regardless of ethnic back-

ground, will learn perfectly whatever language it is exposed to, and

an isolated child prevented from any exposure to language will

learn no language at all. Nevertheless, modern linguists are often

impressed by the striking resemblances among languages all over

the globe. In spite of the obvious and seemingly dramatic differ-

ences among them, linguists are increasingly persuading themselves

that the observed degree of variation in language structures is muchless than we might have guessed in advance, and hence that there

are important universal properties shared by all languages.

In the 1960s, the American linguist Noam Chomsky put forward

a bold hypothesis to explain this apparent universality: according

to his innateness hypothesis, a number of important characteristics

of language are built into our brains at birth, as part of our genetic

endowment, and hence we are born already ‘knowing’ what a

human language can be like. In this view, then, learning a parti-cular language is merely a matter of learning the details which dis-

tinguish that language from other languages, while the universal

properties of languages are already present and need not be

learned.

The innateness hypothesis was controversial from the start, and a

number of critics, among them philosophers and psychologists,

took vigorous issue with Chomsky’s position, arguing that there is

no evidence for innate linguistic knowledge, and that the acquisi-tion of a first language could be satisfactorily explained in terms of

the all-purpose cognitive faculties which the child uses to acquire

other types of knowledge about the world. This controversy

reached a head in 1975, when Chomsky debated the issue with

one of his most distinguished critics, the Swiss psychologist Jean

Piaget.

Chomsky and his supporters have responded in several ways.

First, they attempt to point to identifiable universal properties oflanguage, what they call universal grammar (itself a deeply con-

troversial notion); these properties they claim to be arbitrary,

unexpected and in no way deducible from general cognitive princi-

ples. Second, they point out that children never make certain types

of errors which we might have expected. For example, having

learned The dog is hungry, they can produce They dog looks hungry,

INNATENESS HYPOTHESIS

119

yet, having learned Susie is sleeping, they never produce *Susie

looks sleeping. Third, they invoke the poverty of the stimulus. Bythis term they mean that the data available to the child are

quite inadequate to account for the knowledge which the childeventually acquires. For example, the usual rules of question-

formation in English seem to predict that a statement like The girls

who were throwing snowballs have been punished should have a cor-

responding question *What have the girls who were throwing been

punished? In fact, every English-speaker knows that this is impos-

sible, and no child or adult ever tries to construct such questions.

However, there seems to be no way that this constraint could pos-

sibly be inferred from what the child hears, and Chomsky thereforeinvokes a universal principle, supported by comparable data from

other languages, which he takes as part of our innate linguistic

endowment.

See also: genetic hypothesis of language; universal grammar

Further reading: Aitchison 1989; Piattelli-Palmarini 1979; Sampson1997; Steinberg 1993.

INTEGRATIONALISM

An approach to linguistics based on communicativeness, which

argues that separating langue and parole is a mistake (or at least a

misreading of the work of Ferdinand de Saussure). Integrationalists

argue that the typical separation of language from its context in

what they term segregationalist linguistics produces an object for

analysis that can no longer be considered as language. Instead of

being an ‘add-on’ factor, the pragmatic, psycholinguistic and social

setting of an utterance is an inseparable part of its nature, and mustbe described holistically by the linguist. While persuasive as an

idealization, it is difficult to see what an integrationalist practical

analysis would actually look like, since every analysis would need to

be unique and unrepeatable, and this strikes at the heart of any

claim of linguistics to be scientific, generalisable or replicable.

See also: langue; parole

Further reading: Davis 2001; Harris 1998; Harris and Wolf 1998;Toolan 1996.

INTEGRATIONALISM

120

INTERNAL RECONSTRUCTION

A method in historical linguistics which can be applied to a single

language to recover information about its past. The name internalreconstruction is given to several related but different procedures.

What most of them have in common is this: we observe that a cer-

tain pattern exists in the language of interest, but that certain forms

are exceptions to the pattern; we hypothesize that the exceptional

forms were once regular, and we identify the changes which made

them irregular.

A particularly famous example of internal reconstruction

involves Proto-Indo-European (PIE), the unrecorded ancestor of theIndo-European languages. The young Ferdinand de Saussure

observed that, while the vast majority of PIE verbal roots could be

reconstructed with the general form CeC-, where C represents any

consonant, there were a number with other forms: eC-, aC-, oC-,

Ce:-, Ca:- or Co:-, the last three always with long vowels repre-

sented by a colon. Saussure therefore hypothesized that these

exceptional roots had once been perfectly regular roots of the form

CeC-, but that certain ancestral consonants had been categoricallylost from the language, that some of these vanished consonants had

first altered the quality of a neighbouring vowel e to a or o, and

that all of them had, upon disappearance, induced lengthening of a

preceding vowel. This analysis makes all the exceptional roots the

result of regular phonological change applying to what were ori-

ginally perfectly normal roots. These hypothetical lost consonants

have become known as laryngeals, and Saussure’s laryngealhypothesis was eventually confirmed by the discovery that oneIndo-European language, Hittite, actually preserves some of these

consonants.

See also: comparative reconstruction

Further reading: Crowley 1997; Fox 1995; Hock 1986; Trask 1996.

INTERNATIONAL LANGUAGE

A language which is widely used, for a variety of purposes, by

people in different countries, especially by people for whom it is not

a mother tongue. As a consequence of various political and social

circumstances, a single language may come to be used extensively


121

by people in a number of countries, most of whom speak various

other languages as their mother tongues. In medieval Europe, Latin

was everywhere the language of scholarship, science, diplomacy,

religion and (usually) the law, and people working in these areascould all write (and sometimes speak) Latin. In eighteenth- and

nineteenth-century Europe, French was the international language

of diplomacy, of fine arts and high culture, and of polite society

generally; most educated people could and did speak French, even

if their mother tongue was (say) Russian or German. Chinese and

Arabic have likewise served at times as international languages, in

east Asia and in the Muslim world, respectively.

Today, however, English is beyond dispute the premier interna-tional language throughout the world. English is everywhere the

first language in such domains as business, science, technology,

communications and popular culture. When a Swedish manu-

facturing company negotiates a business deal in Thailand, the

negotiations are conducted in English. When a Brazilian geneticist

publishes his latest research, he publishes in English. When an

Egyptian pilot flies into Moscow, he speaks to the control tower in

English. When the Secretary-General of the United Nations, theChairman of the International Olympic Committee or the Com-

manding Officer of NATO makes an important announcement, he

makes it in English. Even pop groups from Norway, Russia, France

and Japan often sing in English.

This pre-eminence of English is rather recent; it largely dates

from 1945. Though the British Empire had earlier introduced Eng-

lish, as a first or a second language, into huge areas of the world, it

was chiefly the rise of American political and economic power, andthe accompanying spread of American culture (hamburgers, jeans,

Hollywood films, TV shows, rock music) after the Second World

War which made a command of English such an attractive and

valuable asset in the succeeding decades.

Indeed, English has now become the first global language or

world language the planet has ever seen. That is, English now enjoys

some kind of special status in almost every country in the world: as

the sole official language, as a co-official language, or as the desig-nated principal foreign language. Polyglot countries like Nigeria,

India and Singapore use English for most administrative and com-

mercial purposes; countries like Spain, Germany, Greece, Algeria,

Korea and Japan have abandoned French or Chinese as the pre-

ferred foreign language in favour of English; many large German

business firms now require their senior managers to conduct all


122

internal business in English; Dutch universities have come close to

adopting English as the sole language of instruction for all subjects,

and may yet do so. It has been estimated that nearly one quarter of

the Earth’s population – approaching 1.5 billion people – are nowcompetent in English, and the number is growing all the time.

Nothing like this has ever happened before.

See also: power

Further reading: Brutt-Griffler 2002; Crystal 1997, 2003; Jenkins 2003;Wardhaugh 1987.

INTERNATIONAL PHONETIC ALPHABET

An agreed set of symbols for representing speech sounds. The lan-

guages of the world employ a rather large number of distinguish-

able speech sounds, and, for serious work in phonetics and

linguistics, it is essential to have an agreed set of symbols for

representing these sounds as explicitly as possible. The twenty-sixletters (graphemes) of the roman alphabet are not nearly enough for

the task, and consequently various systems have been invented for

the purpose, most of them based upon the roman alphabet supple-

mented by various additional symbols and/or diacritics (‘accents’).

By far the most widely used system is the International PhoneticAlphabet, or IPA. The IPA was invented in 1888 by the Interna-

tional Phonetic Association, the professional body of phoneticians.

The Association has frequently revised and expanded its alphabet,the most recent revision having taken place in 2005: this IPA table

is reproduced at the beginning of this book.

The policy of the Association is to keep the use of diacritics to a

minimum, and hence the IPA employs avery large number of specially

designed characters, such as [S] for the consonant in English shy, [ ]

for the velar nasal in English singing, [ ] for the sound spelled <ll>

in Welsh, as in Llanelli, and [£] for the sound spelled <gl> in Italian,

as in figlio ‘son’. This makes the alphabet cumbersome to print, butnevertheless the IPA is the most widely used phonetic alphabet in

the world; even many good dictionaries of English and of other

languages now often use the IPA for representing pronunciations.

The ordinary roman letters often have obvious values, as with [k],

[s], [m] and [i] (this last is the vowel of Italian si ‘yes’), but a few of

these have values which are unexpected for English-speakers: [j] for


123

the initial consonant of English yes, [y] for the vowel of French tu

‘you’, [x] for the final consonant of German Bach and [r] for the

trill of Spanish perro ‘dog’.

Double characters are used for certain purposes: for example, [tS]for the affricate occurring twice in English church. Diacritics are

used for a wide variety of purposes: for example, [tw] is a labialized

(lip-rounded) version of [t], [tn] is a nasally released version of [t]

(released through the nose, not through the mouth), [b~], is a creaky-

voiced version of [b], and [ i~] is a retracted version of [i] (a vowel

resembling [i] but pronounced a little further back in the mouth).

Length is marked by a special colon, so that [i:] represents a long

version of [i]. There are also symbols for representing various typesof tones, pitches, boundaries and pauses.

Though they are primarily designed for representing speech

sounds (objective physical events), the IPA symbols are naturally also

widely used for representing the phonemes of particular languages.

For example, the initial consonant of English think is phonetically the

dental fricative [h] for most speakers, and so the phoneme realized

in this way is commonly represented as /h/. But note carefully that a

conventional phoneme symbol consisting of an IPA symbol inphoneme slashes may not in fact be pronounced in the way the IPA

symbol would suggest; for example, the phoneme at the beginning of

English red is customarily represented as /r/, for orthographical con-

venience, but probably no native speaker of English ever pronounces

this word with the trill [r]: it’s just that the IPA symbols that repre-

sent precisely what English-speakers say are awkward to print, and

moreover different speakers of English use different pronunciations

of /r/, pronunciations which would be represented by different IPAsymbols (such as [Q], [�], [R], [R] or even [t], for example). An IPA

symbol in square brackets is (or should be) intended to represent a

real speech sound accurately; an IPA symbol in phoneme slashes is

just a convenient way of representing some phoneme in some lan-

guage and may not be a faithful guide to phonetic reality.

In the USA, many people have used a different phonetic alpha-

bet, called American transcription. This uses many different symbols

from the IPA, with diacritics in preference to special characters; itwas designed for people writing on a typewriter. With the increas-

ing use of computers and word processors which can produce IPA

characters, the American system is slowly giving way to the IPA,

though you might still find it in older texts.

See also: consonant; phonetics; speech sound; vowel


124

Further reading: Ball and Rahilly 1999; Katamba 1989; Ladefoged1993; Pullum and Ladusaw 1996.

INTERTEXTUALITY

Connections between texts. The concept of intertextuality was

introduced in the 1960s by the French critic Julia Kristeva, drawing

on work by the Russian critical theorist Mikhail Bakhtin. Most

obviously, the term can be applied to the prominent allusions made

in one literary work to another work, which serve to create coun-

terpoint, continuity, or irony, or draw on authority, or increaseartistic richness or resonance, for example. However, Kristeva sees

every text as constituting an intertext in a succession of texts

already written or yet to be written. A version of this idea has

recently begun to be incorporated into the linguistic analysis of

texts. The general idea is that a text does not exist in isolation and

cannot be fully appreciated in isolation; instead, a full under-

standing of its origins, purposes and form may depend in important

ways on a knowledge of other texts. A sonnet may depend upon thereader’s familiarity with the sonnet-writing tradition; a newspaper

story may depend upon previous news stories; a political speech

may invoke earlier speeches and political statements; even a recipe

may depend upon the reader’s acquaintance with other recipes.

See also: genre; text

Further reading: Fabb 1997; Thibault 1994.

INTONATION

Variation in the pitch of the voice during speech. In science-fiction

films, robots are often made to speak in a dead-level pitch, with no

rises and falls. The (intended) effect is one of inhuman speech, for no

healthy human being ever speaks in such a way. Instead, the pitchof our voice rises and falls in structured ways during each utterance,

and the resulting pattern is the intonation pattern of the utterance.

Every utterance in every language, without exception, is pro-

duced with some intonation pattern imposed on it. These patterns

are largely peculiar to individual languages and accents, though

certain universal tendencies can be observed. For example, it is very

INTONATION

125

common in languages for yes–no questions (as in Are you coming

with us?) to be uttered with a final rise in pitch, while statements (as

in This is my wife) tend more often to be uttered with a final fall –

but Australians and Americans often utter statements with a finalrise, a phenomenon which is highly conspicuous to other speakers

of English, and which has popularly been dubbed ‘uptalk’ or the

more technical high-rising terminal. This feature diffused rapidly

across the English-speaking world in the 1990s, and has even found

its way into other languages.

The intonation of English has been seriously studied since the

1920s, at least. The British tradition has been to analyse intonation

in terms of contours, or tunes, superimposed on sizable chunks ofan utterance; the American tradition, in contrast, prefers to

decompose intonation patterns in terms of jumps between several

discrete levels.

See also: prosody; suprasegmental

Further reading: Cruttenden 1986; Tench 1996; Yip 2002.

INTUITION

Your ‘gut feeling’ about the facts of your native language. We all

have intuitions about our own language: about what is normal,

acceptable, unusual, strange or impossible, or about what a given

form means and when we might use it, if at all. The issue, and it is a

central one, is how much trust (if indeed any at all) we should placein speakers’ intuitions in compiling our descriptions of language.

The American linguists of the first half of the twentieth century,

with their adherence to empiricism, generally preferred to base their

descriptions entirely on the observed spontaneous usage of native

speakers. At times they found it necessary (as we still do today) to

interrogate speakers directly to obtain elusive information: ‘How

would you say X?’; ‘Is Y a possible form?’; ‘What does Z mean?’

But, on the whole, they regarded speakers’ judgements as a lesssecure source of information than spontaneous usage.

In the 1950s, the American linguist Noam Chomsky and his fol-

lowers put forward a dramatic new proposal: that important facts

about a language could be obtained directly from speakers’ intui-

tions. That is, they proposed that you could find out important

things about your own language by merely asking yourself questions

INTUITION

126

like ‘Is construction X possible, and, if so, what does it mean?’

Since almost all the early Chomskyans were speakers of English,

this meant that their descriptions of English were very largely based

upon their own considered opinions about their own usage. Andthis policy has proved to be deeply controversial.

For one thing, it didn’t take long for sceptical linguists to

demonstrate conclusively that native speakers’ intuitions were

sometimes hopelessly wrong. Again and again, the sceptics found

that native speakers of English and of other languages, on being

asked ‘Is construction X possible for you?’, would assert strongly

that X was not possible, and would sometimes further insist that

they couldn’t even understand X – and then, a few minutes later,these same speakers would use X in their own spontaneous speech.

A good example is the north-eastern American locution ‘They’re a

lousy team any more’, which, on interrogation, is sometimes rejec-

ted as ungrammatical and incomprehensible by people who never-

theless use this form freely in their own speech. (The example

means ‘They used to be a good team, but now they’re lousy’.)

For another thing, the linguists who were working on English

often had a theoretical stake in the outcome of questions, and timeand again a linguist pondering the status of a doubtful example

would consult his (or her) own intuitions and find that those intui-

tions gave him just the answer he needed to make his current ana-

lysis work. For example, one linguist, pondering the case of ‘I saw

us in the mirror’, reached the conclusion that this was

ungrammatical – which was exactly the result he needed to make

his new theory work. Yet most speakers of English who are unac-

quainted with the theoretical issue and have no stake in it not onlyagree that this sentence is perfectly grammatical but use such sen-

tences freely in their spontaneous speech.

Consequently, many linguists not of a Chomskyan persuasion

today reject intuitions as a reliable source of information and insist

on verifiable and observable empirical evidence to test their theories.

Much debate concerns methods of minimizing the observer’s paradox,

so that the fieldworkers themselves do not inadvertently alter the

data. In sociolinguistics and applied linguistics, some skilful techni-ques for linguistic data collection have been developed as a result.

See also: generative grammar; Government-and-Binding Theory; socio-linguistics

Further reading: Labov 1975; Sampson 1975.

INTUITION

127

IRREALIS

Irrealis verb forms indicate a grammatical mood in which the sense

expressed is non-actual, doubtful, speculative or is in some otherway not real. For example, If I were rich, I’d buy ten of them con-

tains two irrealis verbs: the second (I would) describes an imaginary

state conditional on being rich; and the first (I were) expresses an

unrealized possibility. Like many Indo-European languages, English

expresses this latter sense using the subjunctive mood form of the

verb (were rather than the indicative mood form was), though the

subjunctive has been falling out of English usage across the world

and is really only maintained in fossilized forms like this. Greek,Sanskrit and Georgian also possess an optative mood for irrealis

senses, which express desired or hoped-for states. Irrealis forms are

used for necessity, obligation, possibility, fears and any sort of

counterfactual. In cognitive linguistics, irrealis forms are particularly

interesting since they cue up parallel or alternate knowledge frames

which the hearer has to monitor.

See also: mood

Further reading: Palmer 2001.

KINSHIP TERMS

The system of terms available in a given language for naming rela-

tives. Kinship terminology varies widely among languages. WhereEnglish has only uncle, many other languages have different words

for ‘father’s brother’ and ‘mother’s brother’. English distinguishes

niece and nephew, but some other languages have only a single word

to cover both, and this same word sometimes covers also ‘grandson’

and ‘granddaughter’. In English, both men and women have only

sisters, but in Basque a man has an arreba while a woman has an

ahizpa, while in Seneca there are different words for ‘older sister’

and ‘younger sister’. German has a useful word geschwister whichmeans ‘sibling’ but sounds more like the formality level of English

‘brother’ or ‘sister’.

The scope for variation is enormous, but anthropological lin-

guists have found that most kinship systems can be analysed into

fairly orderly combinations of a few semantic features, such as

[male/female ego], [male/female referent], [older/younger], [ascending/

IRREALIS

128

descending generation]. Several particular systems are found

to recur widely in the world’s languages, such as the famous

Omaha system found in certain North American languages and

elsewhere.

See also: anthropological linguistics

Further reading: Foley 1997; Wardhaugh 2005.

LANDMARK

A term from cognitive grammar to explain how prepositions are

conceptualized. Prepositions like over, through, into, from, against,

about and so on are metaphorically derived from an idealized spa-

tial concept which is called an image-schema. The notion of over,

for example, involves one object in the focus of attention (the

figure) describing a motion path over the top of an object that

receives less attention (the ground). Figure and ground in image-

schemas are termed trajectors and landmarks respectively.

See also: cognitive linguistics; foregrounding; trajector

Further reading: Croft and Cruse 2004; Langacker 1987–91, Ungererand Schmid 1996.

LANGUAGE

The central object of study in linguistics. The term language covers

several rather different concepts which need to be carefully

distinguished.

To begin with, of course, we need to distinguish between an

individual language – such as English or Swahili – and language in

general. Most linguists believe that all individual languages neces-

sarily possess important properties in common – otherwise, lin-guistics would be a somewhat unrewarding discipline – and every

individual language is therefore a combination of these universal

properties with a number of accidental and often idiosyncratic fea-

tures. For many (not all) linguists, it is these universal properties

which are of greatest interest, but the only way we can get at these

properties is by scrutinizing individual languages.

LANGUAGE

129

In this enterprise, strategies differ. Some linguists prefer to ana-

lyse a few languages in exhaustive detail, in the hope of identifying

subtle abstract principles concealed deep in the data; these principles

constitute universal grammar. Others, though, dismiss this approachas narrow and misleading, and prefer to proceed by surveying large

numbers of structurally different languages and looking for both

generalizations and interesting diversity.

The ultimate goal of linguistics is the elucidation of the human

language faculty (called langage by Saussure). To this end, lin-

guists have usually found it essential to distinguish between the

abstract mental system of rules, principles and constraints which

are shared by speakers (called langue by Saussure and competenceby Chomsky, though the terms are not quite equivalent) and the

real utterances produced by individual speakers on particular

occasions (Saussure’s parole, Chomsky’s performance). More

recently Chomsky has introduced a further distinction: he sug-

gests that an individual language may itself be viewed either as a

set of rules and principles in the minds of speakers (his I-language – intensional, internal and individual – which is the

object of linguistic study), or as a set of possible sentences insociety (his E-language – external – which cannot be studied);

this latter concept is still an abstraction, distinct from actual

utterances.

One way of getting at the abstract properties of natural lan-

guages, championed by Chomsky in his early days and still pursued

by others (though no longer by Chomsky), is to devise formal

grammars which can characterize formal languages. The idea is to

compare the properties of these formal languages with the observedproperties of natural languages to see which kinds of formal gram-

mars give the best fit. This is the motivation behind the construc-

tion of generative grammars.

By complete contrast, others have seen the formalist linguistic

separation of language from its context as a fundamental mistake.

Sociolinguists and applied linguists, for example, regard the collec-

tive performance of language by speech communities as the proper

object of study. Those working in pragmatics or discourse analysissimilarly work with the social and ideological aspects of language in

use. Integrationalists regard the study of language as the study of

communication, and no piece of language can be wrenched from its

interpersonal, social and uniquely occurring situation without fun-

damentally altering the nature of the data. Though formalism con-

tinues to dominate theoretical linguistics, the field of linguistics

LANGUAGE

130

should properly be seen as being at the centre of this rich and

broad-based research movement.

See also: competence; language faculty; natural-language processing;performance

Further reading: Chambers 2003; Crystal 1997; Trask 1995.

LANGUAGE ACQUISITION

The process by which a child acquires its mother tongue. The

acquisition of a first language is arguably the most wonderful feat

we perform in our whole life, and we do it at an age when we can

hardly do anything else. An explanation for this feat is now con-

sidered to be one of the central tasks of linguistics.

The serious linguistic investigation of acquisition dates mainly

from the 1940s, when the Russian linguist Roman Jakobson pub-

lished a pioneering study. In 1957, the American psychologist B.F.Skinner published Verbal Behavior, an attempt at explaining acqui-

sition within the framework of behaviourism, but this book was

savagely reviewed by the American linguist Noam Chomsky, who

argued persuasively that Skinner’s account was hopelessly inade-

quate to explain anything of interest.

From about the 1960s, intensive studies of children acquiring

their first language became increasingly frequent; particularly

influential was Roger Brown’s 1973 book A First Language. Thisbody of work quickly established that early acquisition proceeds

through a sequence of well-defined stages, called cooing, babbling,

the one-word stage and the two-word stage. After this, it becomes

impossible to recognize well-defined stages, though particular con-

structions, such as questions and negation, are found to develop in

a series of well-ordered stages which are highly consistent not only

across children but across languages.

A further crucial observation is that acquisition of a first lan-guage seems to be possible only up to a certain age, the cutoff age,after which it is no longer possible, an observation formalized in

the critical period hypothesis put forward by the neurologist Eric

Lenneberg in the 1960s. Children deprived of exposure to language

during this crucial period are feral children, and several such

unfortunate cases have been studied.

LANGUAGE ACQUISITION

131

We now know a great deal about what children can be observed

to do as they acquire their first language, and attention has turned

increasingly to theoretical interpretation of the raw data. Analysts

have been impressed both by the speed of acquisition and by itsfundamental orderliness. Many, like Chomsky, have been further

impressed by the seeming observation that the data available to the

child from adult utterances seem to be inadequate to account for

the knowledge acquired. This poverty-of-the-stimulus argument led

Chomsky to propose his innateness hypothesis, by which children

are born already knowing what human languages are like, and only

need to acquire the details of the language they are learning; to this

end, Chomsky postulated a specific language acquisition device(LAD) in the human brain.

However, undoubtedly the greatest single advance in the study of

acquisition has been the realization that language acquisition is not

merely a passive affair in which the child soaks up bits of language

that come her way. Instead, it is an active process: children actively

construct their language. There is now no doubt about this: children

take the clues available to them and use these clues to construct

their own grammatical rules, rules which grow in sophistication asacquisition proceeds. This conclusion is strongly confirmed by the

observation of acquisition in unusual circumstances, such as deaf

children acquiring a sign language or children hearing only a pidgin

and quickly turning it into a creole. This determination to acquire

language is very powerful and must surely be part of our biological

endowment; the Canadian psycholinguist Steven Pinker has dubbed

it the language instinct.

See also: bioprogram hypothesis; critical period hypothesis; genetichypothesis of language; innateness hypothesis; language acquisitiondevice; language instinct

Further reading: Berko Gleason 1997; Crystal 1997; Fletcher and Mac-whinney 1995; Goodluck 1991; Ingram 1989; Lust 2006; Owens 1996,1991; Steinberg 1993.

LANGUAGE ACQUISITION DEVICE

A hypothetical mental organ dedicated to the acquisition of a first

language. In the 1960s, the American linguist Noam Chomsky

began developing his innateness hypothesis, by which we are born

LANGUAGE ACQUISITION DEVICE

132

already knowing what human languages are like. Chomsky further

elaborated his hypothesis by arguing that children must possess a

language acquisition device, or LAD, a specific mental organ (a

structure in the brain) which is dedicated to extracting from hap-hazard and often degenerate speech the generalizations required for

the child to construct the necessary rules of phonology and grammar.

Chomsky’s argument was an argument from necessity: because of

the seemingly formidable obstacles to language acquisition

explained in that entry, there simply must be some specialized neu-

rological structure which can extract generalizations and construct

suitable rules, even rules which are not overtly illustrated by the

speech the child hears. But the LAD is purely hypothetical: no onehas yet identified any areas or structures in the brain which seem to

have the required characteristics, and there is no shortage of critics

who see the LAD as a fantasy.

In recent years, Chomsky himself has seemingly abandoned his

claims for the LAD in favour of an even stronger claim: he now

believes that so much information about the nature of human lan-

guage is already present in our brains at birth that all the child has

to do is to ‘set a few switches’ to the correct values for the languagebeing acquired. This is his parameter-setting model, and it too is

deeply controversial.

A contrary view from cognitive linguistics suggests continuities in

humans’ physically embodied condition, with connections between

perceptual and motor skills development out of which language

emerges and is distributed across the brain. However, the question

of where language comes from – both individually and for our

species – remains a difficult and controversial area.

See also: innateness hypothesis; language acquisition; language instinct

Further reading: Aitchison 1989; Lust 2006.

LANGUAGE AREAS

The regions of the brain which are devoted to particular aspects of

the use of language. As part of their investigations into aphasia in

the mid-nineteenth century, the French surgeon Paul Broca and the

German neurologist Carl Wernicke identified two well-defined areas

in the brain which play a crucial role in the use of language. Both

are located in the left hemisphere of the cerebral cortex (the wrinkly

LANGUAGE AREAS

133

grey outer layer of the brain) in the vast majority of people, though

a few people have them on the right side or even on both sides.

Wernicke’s area is located behind and above the ear. It is

responsible for comprehension and also for access to ordinaryvocabulary in speaking. Broca’s area is located close to the temple.

It is responsible for providing the necessary grammatical structure,

including grammatical words and affixes; in speech, it also controls

intonation and the fine muscular movements of the speech organs.

Originally identified by post-mortems carried out on brain-

damaged patients who had suffered from aphasia, these areas can

also be observed to be functioning in normal, healthy, conscious

subjects by the use of modern brain scanners.Since Wernicke’s area is close to the part of the brain which

processes auditory input, and since Broca’s area is close to the area

controlling muscular movements, we require only one further link

to make sense of the whole arrangement. This link, the arcuatefasciculus, was found long ago. It consists of a J-shaped bundle of

fibres connecting Wernicke’s area to Broca’s area.

Consequently, we now have a good picture of the organization of

language in the brain. During listening, speech signals are passedby the ears to the auditory part of the brain, which processes the

sounds and sends the result to Wernicke’s area for interpretation.

During speech, Wernicke’s area provides the ordinary vocabulary,

via the arcuate fasciculus, to Broca’s area, which embeds this voca-

bulary into the required grammatical structure and then sends its

instructions to the organs of speech. With the necessary changes for

switching from a sound medium to a visual medium, essentially the

same things happen during the use of sign language. With a highdegree of reliability, damage to any of these areas produces the

expected language disability (though the brain has proven itself

remarkably adept at routing skills around damaged areas to learn

to talk again).

See also: aphasia; language disability; neurolinguistics

Further reading: Aitchison 1994, 1998; Akmajian et al. 1995; O’Gradyet al. 1996; Steinberg 1993.

LANGUAGE CHANGE

Change in languages over time. One of the fundamental facts about

living languages is that they are always changing. New words, new

LANGUAGE CHANGE

134

pronunciations, new grammatical forms and structures, and new

meanings for existing words are always coming into existence, while

older ones are always dropping out of use. It is absolutely impos-

sible for a living language to avoid changing.The motivations for change are many and various, and only some

of them are reasonably well understood. New objects, new concepts,

new activities all require new names; at the same time, old objects

and activities may cease to exist, and their names may die with

them. Certain linguistic forms may acquire social prestige and

spread to the speech of those who formerly did not use them. The

physiological characteristics of the mouth may tend to favour cer-

tain changes in pronunciation, but such changes may disrupt for-merly regular grammatical patterns, introducing irregularities which

may later be removed in one way or another. Syntactic structures

which come to be frequently used may be reduced to simpler

grammatical forms. And language contact may induce speakers to

import forms and usages from other languages.

Such constant change means that a language at any point in time

is always significantly different from its direct ancestor of some

centuries earlier, and often vastly different from its ancestor of oneor two millennia earlier. Moreover, a language spoken over a sizable

area does not change everywhere in the same way, and so, over

time, it breaks up, first into regional dialects and then, eventually,

into several very different languages, producing a language family.

The diffusion of change happens irregularly and inconsistently, and

it is largely this that is responsible for language variation (studied in

sociolinguistics).

The study of language change itself is historical linguistics, andthis discipline has enjoyed great success in working out the innu-

merable changes which have applied in the past to individual lan-

guages and families; it has also made progress in identifying, and

sometimes explaining, principles of language change: some types of

change, we now know, are more natural, more frequent and more

readily explicable than others.

Not infrequently, speakers take exception to the presence in their

language of certain changes, or even of all changes, and they cam-paign to ‘stamp out’ those innovations of which they particularly

disapprove. Sometimes they even agitate to ‘fix’ their language into

a particular form admired by them, like a dead butterfly in a spe-

cimen box, with no further changes to be tolerated except after

protracted deliberation by suitable authorities. Well, it is true that

certain changes may lead to a (temporary) reduction in the expressive

LANGUAGE CHANGE

135

power of a language (though most do not, and many changes

actually increase its expressive power), and informed commentary

on these matters may be valuable in educational contexts. On the

whole, though, railing against language change is at best a waste oftime and at worst a display of ignorance of the processes which

make language a rich human achievement.

See also: historical linguistics; language contact; language family; languagemyths; prescriptivism

Further reading: Aitchison 2001; Auer et al. 2005; Chambers et al. 2002;Cooper 1989; Crowley 1997; Crystal 1997; Fennell 2000; McMahon1994, 2000; Milroy 1992; Trask 1994, 1995, 1996.

LANGUAGE CONTACT

Changes in one language resulting from the influence of another

language. The speakers of any given language are almost always in

some kind of contact with the speakers of one or more other lan-guages, for any of several reasons. When two different languages are

spoken in adjacent areas, speakers on both sides of the boundary

will be exposed to the other language, and may often gain some

fluency in that other language, creating a dialect chain across poli-

tical national boundaries Because of conquest or migration, speak-

ers of two or more languages may be mixed together in a single

community. Speakers of one language may travel and become

exposed to different languages spoken elsewhere. And, of course, inmodern times the mass media have brought awareness of a number

of languages into regions in which these were formerly unknown.

In all such cases, speakers of one language may, deliberately or

unconsciously, introduce into their language features of another

language to which they have been exposed, and we therefore speak

of language contact, or simply contact.The consequences of contact may range from the trivial to the

far-reaching. At the simplest level, speakers may merely take over afew words from their neighbours; this is called borrowing, and the

words borrowed are loan words in the receiving language. This

happens most readily because the words are the names of genuinely

new things: for example, English-speakers had never seen coffee, or

boomerangs, or tobacco, or chocolate, or pizzas until they encoun-

tered them being used by speakers of various other languages, and

LANGUAGE CONTACT

136

so took them over along with their foreign names. But it can also

happen purely for reasons of prestige: the enormous prestige of

Norman French in England after the Conquest brought thousands

of Norman French words into English, where they often displacedtheir native equivalents (as when face displaced native English

andwlita), or shifted their English counterparts into other meanings

(which accounts for our beef from cattle, our mutton from sheep,

and our pork from pigs).

But contact can go much further than this, affecting grammar

and pronunciation. For example, the Celtic language Breton,

spoken in Brittany, has acquired a French-style uvular /r/, and it

has been losing its native phoneme /h/, absent from French. TheMayan languages of Mexico and Guatemala have acquired a

number of new phonemes from the local prestige language, Span-

ish. Some varieties of Scottish Gaelic have lost the inflected pre-

positions of that language and replaced them with prepositional

phrases comparable to English ones. Among the Semitic languages

of Ethiopia, the original verb–subject–object word order of sen-

tences has been largely changed to the subject–object–verb order

typical of the neighbouring Cushitic languages.Few languages are, or have ever been, sufficiently isolated to

avoid some degree of contact, and hence virtually every language

shows some evidence of ancient or modern contact with other lan-

guages. On occasion, speakers of a given language may react unfa-

vourably to such contact by embracing purism, with variable results.

In recent years, the world dominance of English has led to massive

English influence upon languages from French to Japanese.

In extreme cases, the effects of contact may be so overwhelmingthat one language is abandoned entirely by its speakers in favour of

another, in the process called language death.

See also: language death; linguistic area; loan word; purism

Further reading: Hock 1986; Hock and Joseph 1996; Kouwenberg andSingler 2005; McMahon 1994; Trask 1996.

LANGUAGE DEATH

The disappearance of a language as a mother tongue. A language,

particularly a minority language, may come under enormous pres-

sure from a more prestigious or more widely used language spoken

LANGUAGE DEATH

137

nearby. Native speakers of the language under pressure may find

themselves obliged, not only to learn the local prestige language,

but to use it in an ever-greater number of contexts. Eventually, a

time may come when many children are no longer learning thethreatened language as their mother tongue, or are learning it only

imperfectly. At this point we say the threatened language is mor-ibund or dying, and the almost inevitable result is that, within a

generation or two, no one will be able to speak the threatened lan-

guage at all. This is language death, and the process which leads to

it is language shift.Language death is usually gradual. In any given place, at any

given time, some children are still learning the dying language astheir mother tongue, while others are learning it only imperfectly

and still others not at all. The language may disappear completely

from some areas while surviving in others. Occasionally, a language

dies because its speech community is suddenly wiped out, whether

by natural disaster (such as the 1998 Papua New Guinea earth-

quake which destroyed the populations speaking Arup, Malol, Sis-

sano and Warapu) or by invasion and genocide (such as the

eradication of native Caribbean languages by European colonizers,or the destruction of Moriori speakers in the Chatham Islands by

invading Maoris from New Zealand over the nineteenth century).

This form of language death is commonly called linguicide.The language spoken by the last generation or so of native

speakers may be very much changed from the language spoken by

an earlier generation. Irregularities may be lost; the more complex

or less frequent forms and sentence patterns may drop out of use;

native words may be massively replaced by words taken from theprestige language; the pronunciation may change so as to become

more similar to that of the prestige language; stylistic variation may

be lost, leaving only a single unvarying style. The final outcome, of

course, is a dead language. Later on, efforts may be made to rein-

vent and codify the dead language, and even resurrect it with native

speakers: modern Greek Katharevusa, modern written Welsh,

Manx and Cornish have all been reconstructed with varying degrees

of success.

See also: dead language; power; language contact; minority language

Further reading: Crystal 2000; Dalby 2003; Hock and Joseph 1996;Krauss 1992; Llamas et al. 2007; McMahon 1994; Nettle and Romaine2000.

LANGUAGE DEATH

138

LANGUAGE DISABILITY

Any pathological condition which has adverse consequences for the

sufferer’s ability to use language normally. Language disabilitiesmust be carefully distinguished from speech defects like lisping and

inability to produce a tapped /r/; these are purely mechanical pro-

blems with the nerves and muscles controlling the organs of speech,

and they have no effect upon the language faculty itself (though

stammering, or stuttering in the USA, does seem to involve some

cognitive interference as well).

A true disability results either from a genetic defect or from

damage to the language areas in the brain. Since the possible typesof damage and defects are virtually limitless, individual sufferers

naturally exhibit an enormous range of disabilities: we hardly ever

find two individuals with exactly the same symptoms. Nevertheless,

it has proved possible to identify a number of fairly specific dis-

abilities and, in many cases, to associate these with particular

genetic defects or with damage to particular areas of the brain.

The best-known disabilities are the several types of aphasia, all of

which result from injury to more or less identifiable areas of thebrain. But other types exist. For example, the Williams syndrome is

known to result from a genetic defect in Chromosome No. 11; this

defect causes both some highly specific physiological abnormalities

and some rather consistent abnormalities in language use. Less well

understood is Specific Language Impairment, or SLI, which devas-

tates the ability to use the grammatically inflected forms of words

correctly (such as take, takes, took, taking) but which has only a few

non-linguistic consequences; the Canadian-based linguist MyrnaGopnik has recently uncovered evidence that this impairment too

may result from a specific genetic disorder, though one which has

not yet been identified.

But the known range of disabilities is positively startling. A

bilingual sufferer may lose one language completely but retain the

second perfectly, and then, after some time, may lose the second

but regain the first. Many types of disability are characterized by

greater or lesser degrees of anomia, the inability to remember wordsfor things. But some sufferers exhibit astoundingly specific deficits,

such as a total inability to use or understand words denoting fruits

and vegetables, with no other problems at all. Some people lose

verbs but retain nouns, and so they have no trouble with the noun

milk (as in a glass of milk), but cannot handle the verb milk (as in to

milk a cow). Some elderly sufferers from Alzheimer’s disease find

LANGUAGE DISABILITY

139

themselves reverting to languages, dialects and accents that they

spoke in their youth, and lose abilities that they gained in later life.

In spite of some impressive progress, we are still very far from

understanding most types of language disability.

See also: aphasia; language areas; neurolinguistics

Further reading: Caplan 1992; Field 2003; Ingram 2007; Pinker 1994;Steinberg 1993; Trask 1995.

LANGUAGE FACULTY

Our biological ability to use language. Human beings are the only

creatures on earth that use language, and many linguists and others

have concluded that we must therefore have some kind of specific

biological endowment for language, one which is totally absent, or

nearly so, from all other living species: our language faculty (the

Swiss linguist Ferdinand de Saussure used the term langage for this,but his label is now little used).

To be sure, this conclusion has been challenged from two direc-

tions. On the one hand, some experimenters have attempted to

teach other species, usually apes, to use some simplified version of a

human language (most often a version of a sign language) and, in

spite of serious problems with their methodology and interpreta-

tions, a few observers are now prepared to accept that these crea-

tures do indeed exhibit a (severely limited) capacity for using

language – though critics of this conclusion are numerous and vig-orous. On the other hand, psychologists like Jean Piaget and

Jerome Bruner have argued that our language faculty, while admit-

tedly real, is not at all an individual and distinctive part of our

biological inheritance, but merely one more manifestation of our

general all-purpose cognitive abilities.

Nevertheless, the majority view among linguists at present is that

our language faculty is real, that it is at least largely distinct from

all of our other cognitive abilities, and that it must be the biologicalresult of some kind of distinctive evolution within the brains of our

ancestors. This is the belief that underlies a number of celebrated

attempts at giving an account of our language-using abilities,

including the genetic hypothesis of language, Chomsky’s innateness

hypothesis, Bickerton’s bioprogram hypothesis, and even the search

for universal grammar.

LANGUAGE FACULTY

140

A constant theme in these investigations is the issue of mod-ularity. Chomsky and others have long argued that our language

faculty must consist of a number of specialized and largely inde-

pendent subcomponents which interact in specific ways to produceour overall linguistic behaviour. More recently, however, some

people have begun to question whether our language faculty as a

whole should itself be regarded as a distinctive part of our mental

equipment. They suggest instead that various aspects of language

use may have entirely separate evolutionary origins, and that what

we call our language faculty is probably an epiphenomenon; that is,

a purely superficial unity which in fact results from the interaction

of diverse structures and processes within our brains, many ofwhich are in no way confined to language behaviour. These debates

will doubtless continue for some time.

The study of all the biological aspects of our language faculty is

sometimes called biolinguistics.

See also: animal communication; autonomy; bioprogram hypothesis;genetic hypothesis of language; innateness hypothesis; origin and evolutionof language

Further reading: Jackendoff 1993; Pinker 1994; Sampson 1997; Steinberg1993.

LANGUAGE FAMILY

A group of languages all of which share a common ancestor. Everyliving language is constantly changing, and the changes which affect

the language in one place do not necessarily affect it in other places.

Consequently, if the language is spoken over any significant area,

then, over time, it will tend to break up into rather distinct vari-

eties. At first these are merely the regional dialects of the language,

but, given sufficient time, these dialects may become so different

from one another that we are forced to regard them as separate

languages: the daughter languages of that single ancestral languageor mother language.

The daughter languages may in turn break up into further

daughter languages, eventually producing a sizable number of lan-

guages all of which started out as nothing more than regional

varieties of that single ancestor. All the languages which share a

common ancestor in this way constitute a single language family,

LANGUAGE FAMILY

141

and we say that all of them are genetically related. Around 300 such

language families have been identified by linguists, some of them

very large, others quite small.

English, for example, is fairly closely related to a group of otherlanguages: Afrikaans, Danish, Dutch, Faroese, Frisian, German,

Icelandic, Norwegian, Swedish, Yiddish and the extinct Gothic; we

call these the Germanic languages, and we believe they are all des-

cended from a single ancestor spoken perhaps around 500 BC, very

likely in Scandinavia. But the Germanic languages in turn share a

more remote common ancestor with a vast number of other lan-

guages, including Welsh, Spanish, Greek, Russian, Armenian, Per-

sian and Bengali (to name just a few), and hence all of theselanguages belong to a single huge family called the Indo-Europeanfamily.

See also: historical linguistics; Indo-European; language change

Further reading: Comrie et al. 1997; McMahon 1994; Trask 1996.

LANGUAGE INSTINCT

The powerful tendency of children to acquire language. Any physi-

cally normal child who is adequately exposed to a language will

learn it perfectly, and a child exposed to two or three languages will

learn all of them. A hearing child normally learns the surrounding

spoken language. A deaf child exposed to a sign language will learn

that. Children exposed only to a pidgin will turn that pidgin into afull language: a creole. A group of children exposed to no language

at all will invent their own and use it.

For a long time, linguists were slow to appreciate the significance

of these observations, and both creoles and sign languages were

widely regarded as peripheral and even inconsequential phenomena

scarcely deserving of serious study by linguists. But times have

changed. Very gradually at first, and then, from about the 1970s,

almost explosively, the examination of these phenomena convincedalmost all working linguists that creoles and sign languages were

every bit as central to the discipline as spoken languages with a

long history.

We now realize that children are born with a powerful biological

drive to learn language, and that only shattering disability or inhu-

man cruelty can prevent a child from acquiring language by one

LANGUAGE INSTINCT

142

means or another. In the 1990s, the Canadian psycholinguist Steven

Pinker coined the felicitous term language instinct to denote this

remarkable aspect of our biological endowment. Our language

faculty, we now strongly suspect, is built into our genes, and learn-ing a first language may not be so different from learning to see: at

birth, our visual apparatus is not working properly, and it requires

some exposure to the visible world before normal vision is

acquired.

Not infrequently, the term language instinct is applied more spe-

cifically to the genetic hypothesis of language and/or to the related

innateness hypothesis.

See also: creole; critical period hypothesis; genetic hypothesis of language;innateness hypothesis; language acquisition; sign language

Further reading: Aitchison 1989; Pinker 1994.

LANGUAGE MYTHS

Widely held but false beliefs about individual languages or about

language in general. For one reason or another, various absurdly

false beliefs about languages have at times taken hold in the popu-

lar imagination and have often proved difficult to stamp out. Some

of these misconceptions have even been maintained by knowledge-

able linguists. Here are a few of these myths.

Primitive languages

Linguists as recent and eminent as Otto Jespersen have maintained

that some languages spoken today are ‘primitive’ or ‘savage’ lan-

guages, characterized by strange grammatical systems supposedly

reflecting a primordial human view of the world. Non-linguists have

at times gone further and supposed that some languages have little

or nothing in the way of grammar, and that they have tiny voca-

bularies lacking abstractions or generalizations and supplementedby grunts and gestures. All this is nonsense: every human language

ever discovered has a rich and complex grammar and a vocabulary

of many thousands of words, and is perfectly adequate for expres-

sing anything its speakers want to express. When they need to express

anything new, they borrow from other languages or innovate in

their own language.

LANGUAGE MYTHS

143

Stadialism

Many linguists formerly believed that every human language

necessarily develops over time from a ‘primitive’ type through

increasingly sophisticated types until finally (perhaps) reaching the

most ‘advanced’ type possible. This idea was promulgated by Eur-

opean linguists who were inclined to see their European languages

as representing the pinnacle of linguistic perfection, with sig-nificantly different languages, and especially those spoken by var-

iously coloured folks, as necessarily representing some earlier stage

of development. This blatantly racist idea is thankfully dead among

linguists, though avowedly non-racist versions of it are still main-

tained in some quarters, and non-linguists sometimes still believe it

is true. It is not. Languages admittedly differ markedly in their

grammatical systems, and they can and do change those systems

dramatically over time; however, there is no tendency for languagesto change in one direction rather than another, and there is nothing

special about European languages.

Those Eskimo words for ‘snow’

By a comical series of events, the legend has grown up that the Eskimo

(Inuit) languages have vast numbers of words for different kinds of

snow. In fact, the several dialects of the two Eskimo languages

variously exhibit between two and four distinct words for snow. This

is about the same as English, with its snow, slush, sleet, blizzard (not tomention skiers’ terms like hard-pack, powder and crust). However, it

is generally true that sets of concepts that are overlexicalized and

underlexicalized in a language (in comparison with other languages)

can give clues as to cultural preoccupations and worldviews.

Basque

The Basque language of western Europe has no known relatives

and it is noticeably different from other European languages, par-

ticularly from those (the great majority) which are related within

the Indo-European family. Consequently, a number of myths havegrown up about it: Basque is the most complex language on the

planet; no outsider has ever managed to learn it; all the verbs are

passive; and so on. In fact, of course, Basque is a perfectly unre-

markable language. Its grammar is highly regular, and it is easy to

learn: today thousands of people speak it as a second language. Its

sentence structure is nearly identical to that of Japanese or Turkish.

LANGUAGE MYTHS

144

It does have a morphology which is somewhat unusual in Europe,

but similar systems are found in hundreds of languages outside of

Europe. Of course, Basque people themselves are partly responsible

for promoting this view of their own separateness and distinctive-ness, with a nationalist political agenda.

The land that time forgot

Somewhere, runs the story, in the Ozarks, or in the Appalachians,or in Derbyshire in England, there’s a village where the locals still

speak perfect Elizabethan English, untouched by the vast changes

which have transformed English everywhere else. No, there isn’t:

this is pure fantasy. There is no such thing as a living language

which doesn’t change. This myth crops up because people occa-

sionally notice that the local English in some corner of the world

preserves one or two old forms which have disappeared elsewhere.

(For example, Appalachian English preserves the a’doing form, as inI was a’shootin’ at some squirrels; this was once universal in English

but has been lost everywhere else.) But every variety of English

preserves a few forms lost in other varieties, and every variety also

exhibits a few innovations not found elsewhere. (For example,

Appalachian English has undergone a change in its vowels such

that Appalachian think sounds to the rest of us rather like thank.)

Similar myths have been maintained by speakers of other lan-guages. Until the eighteenth century, even some linguists believed

that the ancestral language of all humankind was still spoken, in its

pristine state, in some favoured corner of the world; much ink was

spilt over deciding which corner this might be. (For example, one

such linguist argued for the Netherlands, and claimed that Dutch

was the uncorrupted ancestral tongue of all humans. He was

Dutch, of course.) But all languages that are spoken change inces-

santly, and no language anywhere is closer than any other to theremote origins of human speech.

There’s a moral here: don’t believe everything you read. Many

journalists, authors of popular books, and especially website writers

are ignorant of the facts, and they tend to perpetuate eye-catching

falsehoods in place of the more humdrum truth. Your critical sense

should even extend to this book, of course!

See also: folk linguistics; prescriptivism

Further reading: Crystal 1997; Pullum 1991.

LANGUAGE MYTHS

145

LANGUAGE PLANNING

Making deliberate decisions about the form of a language. Very

commonly, a language ‘just grows’: it develops and changes inresponse to countless small decisions made more or less uncon-

sciously by its speakers. But it is perfectly possible, and in some

circumstances necessary, for the future of a language to be deter-

mined in important respects by deliberate, self-conscious decisions,

often made on an official basis. This is language planning, some-

times also called linguistic engineering.Consider the case of Finnish. For centuries, Finland had been a

province of Sweden or of Russia; either Swedish or Russian hadtherefore been the official language of the country, and Finnish

had remained merely the everyday language of most of the popu-

lation. But, when Finland finally achieved independence in 1918,

the Finns naturally chose to make Finnish their new national

language.

But this decision required a great deal of work. First, there was

no agreed standard form of Finnish: instead, there were only innu-

merable local varieties of it, differing in vocabulary, pronunciationand grammar. Second, since Finnish had never been used for such

purposes as law, administration, science, technology and scholar-

ship, there was a great dearth of technical vocabulary in all these

fields: there simply were no Finnish equivalents for terms like man-

slaughter, ministry, molecule, piston and linguistics.

The Finns were therefore obliged to put these things right, and

the task was carried out centrally, by bodies set up by the new

Finnish Government. Specialists in Finnish language drew up anew standard form of the language, and specialists in a huge variety

of disciplines met to agree on suitable Finnish technical terms in

their fields. Their decisions were official, and they were imposed

upon the Finnish taught in schools, published in books and maga-

zines, and broadcast over the radio.

The task was completed satisfactorily, and today there exists a

single standard form of Finnish, known and used by everyone,

together with an adequate and uniform set of technical terms. Butthe work is never done, since new technical terms come into use

every year, in fields ranging from physics to linguistics, most of

them coined in English, and the Finns are constantly obliged to

keep finding Finnish equivalents.

Language planning also includes the deliberate attempt to avoid

sexist, homophobic or other discriminatory discourse (which is

LANGUAGE PLANNING

146

often then characterized by its reactionary opponents as ‘political

correctness’).

See also: national language; official language; sexist language; standardlanguage

Further reading: Bex and Watts 1999; Cooper 1989; Crystal 1997;Holmes 1992; Trask 1996; Wright 2004.

LANGUAGE PROCESSING

The mental activities involved in producing and comprehending

language. Whenever we produce an utterance, or whenever we hear

and understand one, there is a great deal of elaborate activity going

on in our brains. This activity is language processing.Psycholinguists have developed a battery of techniques for work-

ing out the stages involved in language processing. Subjects can be

tested in a laboratory, to find out how the performance of linguistic

tasks is affected by varying conditions, and evidence can beobtained from speech errors. The results, however, are not always

easy to interpret. Nevertheless, linguists have enjoyed some success

in working out the several distinct stages involved in speech plan-ning, the mental processes which allow us to produce utterances.

Comprehension is more difficult: a number of perceptual strategies

have been proposed, and these are moderately successful at

accounting for the understanding of simple utterances, but the

comprehension of more complex utterances is still something of amystery. Recently there have been several hypotheses advanced

within cognitive linguistics, but most of these have yet to be

empirically verified.

See also: perceptual strategy; psycholinguistics; text world theory

Further reading: Aitchison 1989; Akmajian et al. 1995; O’Grady et al.1996.

LANGUE

The abstract language system shared by the speakers of a language.

The term langue was introduced by the Swiss linguist Ferdinand de

LANGUE

147

Saussure in the early twentieth century; in Saussure’s treatment, this

term contrasts specifically with parole, actual utterances. A distinc-

tion along these lines has often been considered essential in lin-

guistics since Saussure’s day, and Noam Chomsky’s more recentdistinction between competence and performance is broadly analo-

gous, but note that Saussure’s langue is the property of a whole

community of speakers, while Chomsky’s competence is the prop-

erty of a single speaker.

See also: competence; language; parole

Further reading: Culler 1986; Sampson 1980.

LATERALIZATION

Specialization between the two hemispheres (halves) of the brain.

By a quirk of evolution, the left side of the brain controls the right

side of the body and vice versa. But there are also notable differ-

ences in the responsibilities of the two hemispheres: the left side ischiefly responsible for analysis (breaking complex things up into

smaller parts), and it handles things like doing arithmetic, solving

equations and determining chronological sequences. The right

hemisphere, in great contrast, is responsible for synthesis (combin-

ing pieces into integrated wholes), and hence it handles things like

recognition and association, and also the enjoyment of music

(though trained musicians learn to use their left hemispheres for

this purpose as well).In the vast majority of people, the language areas of the brain are

in the left hemisphere, though a few people have them on the right

or, rarely, even on both sides.

See also: language areas

Further reading: Akmajian et al. 1995; Ingram 2007; O’Grady et al.1996; Steinberg 1993.

LEMMA

The least marked form of a word, sometimes called the citationform since it is the version most likely to be given as the indexed

LATERALIZATION

148

word in dictionaries. For example, book would be the lemmatized

form of the set of words including books, rebook, booked, bookish,

booker and booking. In relatively low inflected languages like Eng-

lish, it is usually easy to determine the lemma of a set of words(though consider mouse with mice, or go and went, or am, are, is

and be, for example). In highly inflected languages, the situation is

not so easy: Czech, for example, has various forms of dobro ‘good’

(dobre rano ‘good morning’, dobry den ‘good day’, dobry vecer

‘good evening’, dobrou noc ‘good night’ and even mam se dobre ‘I’m

good/fine’), such that selecting one is not straightforward. Or given

a phrasal set, such as pick yourself up, pick myself up, pick himself

up and so on, most people would give pick oneself up as the lemma,even though it would be the least used. Lemmas (lemmata to be

pedantic) are particularly useful when setting concordance pro-

grams to search through linguistic corpuses (corpora).

See also: computational linguistics; lexicon

Further reading: Adolphs 2006; Sinclair 1991.

LEXICAL-FUNCTIONAL GRAMMAR (LFG)

LFG was developed in the late 1970s by the American linguists

Joan Bresnan and Ronald Kaplan. It differs from some other

theories of grammar in holding that the syntactic structure of a

sentence is something more than just the familiar constituent

structure represented by a tree diagram. In LFG, the structure ofa sentence consists of two distinct formal objects: C-structure of

the familiar kind plus a functional structure (or F-structure) whichdisplays certain additional kinds of information. Most important

in the F-structure is the labelling of grammatical relations like

subject and object (these are called grammatical functions in

LFG).

The first part of the name reflects the fact that a great deal of

work is done by the lexical entries, the ‘dictionary’ part of the fra-mework. Lexical entries are usually rich and elaborate, and each

one inflected from a lexical item (such as write, writes, wrote, written

and writing) has its own lexical entry. Lexical entries are responsible

for dealing with many relations and processes handled by different

machinery in other frameworks; an example is the voice contrast

between actives and passives.

LEXICAL-FUNCTIONAL GRAMMAR (LFG)

149

LFG was designed to be convenient for work in computational

linguistics and has found applications there. It was also designed in

the hope that it would prove to be psychologically realistic, but

overt psycholinguistic support has not been forthcoming.

See also: generative grammar; Government-and-Binding Theory; phrase-structure grammar

Further reading: Horrocks 1987; Kaplan and Bresnan 1982; Sells 1985.

LEXICOGRAPHY

The writing of dictionaries. Dictionaries of a sort have been around

for a long time, but ancient and medieval efforts were mostly very

different from our modern ones. To start with, the early ones were

usually bilingual dictionaries, glossaries offering translations of

words from one language into another. The medieval period saw

the production of monoglot works, but these were not usually

arranged alphabetically: instead, the words were grouped by mean-ing (words pertaining to farming, names of fruits, and so on). The

first alphabetical dictionaries of English were not complete: instead,

they were compendia of ‘hard words’, that is, of obscure and diffi-

cult words, often mainly of Latin origin.

By the eighteenth century, books that we can easily recognize as

dictionaries of English were beginning to appear. By far the most

prominent of these in Britain was Dr Samuel Johnson’s great work

published in 1755; Johnson’s American counterpart, Noah Webster,published the first edition of his American dictionary in 1828.

Other dictionaries followed, and lexicography became a recognized

profession in the English-speaking countries.

In the second half of the nineteenth century, an enormous project

was undertaken in Britain: the preparation of a huge dictionary of

English recording every word, every spelling, and every sense

attested in writing in English since the year 1000. Directed by the

Scottish scholar James Murray, this project was carried out bymethods that seem comically primitive today: an army of con-

tributors noted down examples of words and sent them in on slips

of paper, and Murray and his assistants simply built the dictionary

out of these mountains of paper slips. The completed work was

published in a series of volumes between 1884 and 1928, under the

title New English Dictionary on Historical Principles. In 1933 it was

LEXICOGRAPHY

150

republished, with a supplement, as the Oxford English Dictionary,

or OED. Further supplements followed, and finally an expanded

and updated second edition was prepared and published in 1989.

Both editions are available on CD-ROM, making the OED anenormously valuable research tool. A third edition is in process.

Throughout the twentieth century, English dictionaries have been

published in numbers. In the last couple of decades, lexicography

has been revolutionized by the introduction of corpus-based tech-

niques, and modern dictionaries are now usually based upon huge

corpora of English, from which words, forms, spellings, meanings

and grammatical behaviour are extracted, thus allowing lexico-

graphers to appeal directly to the observed facts of usage. Alsonoteworthy are the numerous innovations introduced by the

COBUILD dictionaries in Britain and by other dictionaries written

especially for advanced foreign learners of English. These describe

words as they are observably used by the speech community, rather

than as defined by the lexicographer alone or in prestigious texts.

See also: computational linguistics

Further reading: Green 1996; Ilson 1986; Jackson 2000; Landau 1984;McArthur 1986.

LEXICON

The vocabulary of a language. Every speaker of a language pos-

sesses a certain vocabulary, and this may be divided into his (or her)

active vocabulary, the words which he uses himself, and his passivevocabulary, the words which he understands but doesn’t normally

use. In linguistics, however, we don’t normally speak of the voca-

bulary of a particular language; instead, we speak of the lexicon,the total store of words available to speakers.

Very commonly, the lexicon is not regarded merely as a long list

of words. Rather, we conceive of the lexicon as a set of lexical

resources, including the morphemes of the language, plus the pro-

cesses available in the language for constructing words from thoseresources. For example, given the existence of English verbs like

varnish and scratch, and of the word-forming affixes -able and un-,

it is perfectly possible for you to create the words varnishable and

unscratchable whenever you need them, and to expect to be understood

at once, even if you and the person you are speaking to have never

encountered these words before.

LEXICON

151

Quite apart from the lexicon of a language as a whole, psycho-

linguists are interested in the mental lexicon, the words and lexical

resources stored in an individual brain. Evidence from a variety of

sources, including language disability, has provided a great deal ofinformation about this. There is good evidence that words of a

single grammatical class, and also words of closely related mean-

ings, are stored in the brain ‘in the same place’, whatever that

means exactly; and it is perfectly clear also that words are not

stored in isolation, but are instead stored with innumerable links to

other words which are related in function, which are related in

meaning, which have similar sounds, or which were even learned at

the same experiential moment.

See also: morphology; part of speech; word-formation

Further reading: Aitchison 1994; Bauer 1998; Davis 2001; Hoey 2005;Katamba 1994; Schmitt and McCarthy 1997.

LINGUA FRANCA

A language which is widely used in some region for communication

among people speaking a variety of languages. The original linguafranca (the name means ‘Frankish language’, though the sense is

‘European language’ or ‘Christian language’) was a variety of Ita-

lian, strongly laced with words from French, Spanish, Greek,

Arabic and Turkish, used as a trade language in the eastern Medi-

terranean during the late Middle Ages. Since then, we have appliedthis label to any language which enjoys wide use among speakers of

a variety of different languages in some region.

In the past, this term was applied very broadly, and many of the

speech varieties called lingue franchi (this is the plural if you want

to be pedantic) were pidgins or creoles. Today, though, we more

commonly use the term for a language which is the long-established

mother tongue of some influential group of speakers but which is

none the less widely used for inter-group communication by speak-ers of several other languages, such as Swahili in East Africa,

Hausa in West Africa, or English in Singapore.

See also: creole; international language; pidgin

Further reading: Holmes 1992; Kouwenberg and Singler 2005.

LINGUA FRANCA

152

LINGUISTIC AREA

A geographical region in which several unrelated or distantly rela-

ted languages have striking characteristics in common. With only a

very few exceptions, speakers of a language are always in contact

with neighbours speaking different languages, and they have deal-ings with those neighbours. The resulting language contact means

that words, speech sounds and even grammatical forms may pass

from some languages into neighbouring languages. This is exceed-

ingly common.

In certain cases, however, this contact may be so intense that a

number of striking characteristics diffuse throughout a geo-

graphical region, becoming prominent in a number of languages

which are unrelated or only distantly related. As a result, the lan-guages in question may, in some respects, undergo convergence:they come to resemble one another more closely than they resem-

ble their closest linguistic relatives in other regions. This state of

affairs we call a linguistic area or, using the German term, a

Sprachbund.Several linguistic areas have been identified by linguists, among

them south-east Asia and the Balkans. South-east Asia is a parti-

cularly striking case. Many languages in this area – including Bur-mese, Thai and Vietnamese – are so similar in their structures that

linguists believed for a while that they must all be related in a single

family. But careful investigation has revealed that they are not dis-

coverably related to each other at all; they all have true relatives

elsewhere which are not very similar to them, and the striking

resemblances result purely from convergence among neighbouring

but unrelated languages.

In the Balkans, Greek and Albanian constitute distinct branchesof the Indo-European family, while Bulgarian, Macedonian and

Serbo-Croatian belong to the distinct Slavic branch of Indo-

European, and Turkish is not Indo-European at all. Yet all these

languages participate to varying degrees in a language area: they

share with one another a range of grammatical characteristics

which they do not share with their closest relatives elsewhere.

The study of language areas is called areal linguistics.

See also: language contact

Further reading: Hock and Joseph 1996; Trask 1996.

LINGUISTIC AREA

153

LINGUISTIC RELATIVITY HYPOTHESIS

The hypothesis that the structure of our language to some extent

determines the way we perceive the world. For centuries, scholarshave speculated on possible links between language on one hand

and mind, perception and culture on the other; a prominent exam-

ple is the eighteenth-century German linguist and philosopher

Wilhelm von Humboldt.

Around the beginning of the twentieth century, linguistics began

to emerge from anthropology as a distinct discipline in the USA,

and American linguists, with their anthropological background,

often took a keen interest in the links just mentioned. The Amer-ican linguist Edward Sapir was particularly fascinated by possible

connections between language and thought; he once wrote: ‘Lan-

guage and our thought-grooves are . . . , in a sense, one and the

same’. But it was Sapir’s student, Benjamin Lee Whorf, who was to

develop this idea into its most dramatic form.

Examining several Native American languages, Whorf was struck

by the observation that different languages appeared to divide up

the world differently, and that, further, a concept that was repre-sented as a ‘thing’ in one language might be represented as an

‘event’ or a ‘process’ in another. These purely linguistic differences

he was inclined to see as representing genuine differences in the way

that speakers of different languages perceive the world. Though he

himself never made such an explicit statement, his work soon came

to be interpreted as supporting a remarkable conclusion: the struc-

ture of our language in large measure determines the way we per-

ceive the world. This dictum is known as the linguistic relativityhypothesis, or equally as the Sapir-Whorf hypothesis.

The idea has been controversial from the start. In the 1960s, the

anthropologists Brent Berlin and Paul Kay published a famous

study of basic colour terms in a number of languages, concluding

that, even though languages differ markedly in their linguistic sys-

tems for naming colours, there were important universal character-

istics of colour terms, suggesting that universals of perception

underlay the linguistic differences. But their conclusions were chal-lenged on a number of grounds, most famously by the psychologists

John Lucy and Richard Shweder, who in the 1970s reported some

interesting differences in the behaviour of speakers of different

languages.

Since then any number of linguists, anthropologists and psychol-

ogists have chipped in with fascinating data supporting one side or

LINGUISTIC RELATIVITY HYPOTHESIS

154

the other, but at present there is still no consensus on the degree of

validity which can be reasonably assigned to this hypothesis. How-

ever, psycholinguists have managed to demonstrate that, in memory

tests, it is easier to remember things when we have explicit namesfor them. So, for example, it is easier to remember colours accu-

rately if our language provides a wide range of specific colour

terms. It may be that a weak version of the Sapir-Whorf hypothesis

works, such that linguistic habits encourage a tendency to think in a

particular customary way, but the language itself does not constrain

thought.

See also: anthropological linguistics

Further reading: Berlin and Kay 1969; Duranti 1997; Gumperz andLevinson 1996; Lucy 1992; Lucy and Shweder 1979; Steinberg 1993;Whorf 1956; Wierzbicka 1996.

LINGUISTIC SIGN

A linguistic object possessing both form and meaning. The concept

of the linguistic sign was introduced in the early twentieth century

by the Swiss linguist Ferdinand de Saussure, in whose system it

plays a central role. This concept is a very simple one: every lin-

guistic object has two aspects, or facets: a linguistic form (called by

Saussure the signifiant, or ‘signifier’) and a meaning (the signifie, or‘thing signified’). For example, the English word dog has a parti-

cular form (a sequence of three meaningless phonemes) and also aparticular meaning (a specific kind of animal). The two together

make up a single linguistic sign in English.

Saussure drew particular attention to the arbitrariness of each

linguistic sign: there is no reason why any particular linguistic form

should be associated with any particular meaning, and the pairing

that exists in each case is an arbitrary one.

The recognition of linguistic signs lies at the heart of our view of

any given language as a symbolic system. Saussure’s idea oflinguistic signs was also influential in shaping the discipline of

semiotics.

See also: arbitrariness; iconicity; semiotics; symbolic system

Further reading: Culler 1986; Peirce 1991.

LINGUISTIC SIGN

155

LINGUISTICS

The scientific study of language. There were significant traditions of

language study in ancient India, in ancient China, in ancient Greeceand Rome, among the medieval Arabs and Jews, and elsewhere.

Most of these investigations, though, were solely confined to

studying the local prestige language. Modern linguistics does not

derive from these older traditions; instead, it grew up from fresh

beginnings in Europe and the USA.

By the seventeenth century, a few European scholars and philo-

sophers were beginning to interest themselves in general questions

about the nature of language, and between the seventeenth andnineteenth century, scholars like Descartes, Locke and Humboldt

made a number of significant contributions. But, with a few excep-

tions, these men typically knew nothing about any languages other

than the major languages of Europe, and their work suffered from

a lack of data, with the result that much of it was speculative and a

priori.

By the end of the eighteenth century, historical linguistics had

begun to be firmly established, and throughout the nineteenth cen-tury the historical study of language was for many people synon-

ymous with the scientific study of language. Towards the end of the

century, though, a number of linguists began turning their attention to

the serious study of the structure of language from a non-historical

point of view. Prominent among them were von der Gabelentz,

Kruszewski and Baudouin de Courtenay. But the most influential

figure, by far, was the Swiss linguist Ferdinand de Saussure.

Though he had been trained as a historical linguist, and thoughhe had made major contributions to historical studies, Saussure

began to focus on some more general questions of language struc-

ture and to reach some profound conclusions. He failed to publish

this work, but, after his death, his former colleagues and students

edited his lecture notes into a book, which was published under

Saussure’s name in 1916: this book is the famous Cours. So great

has been the influence of the book that Saussure has been dubbed

‘the father of linguistics’. The European linguistic tradition, with itsheavy theoretical bias, largely derives from Saussure’s work, though

in Britain Bronislaw Malinowski and (especially) J.R. Firth inde-

pendently developed a more strongly data-oriented descriptive

approach born of anthropological fieldwork.

Meanwhile, in the USA, anthropologists were undertaking the

study of the dying Native American languages. This study was

LINGUISTICS

156

keenly promoted by Franz Boas, who is often regarded as the

founder of the American linguistic tradition; Boas’s successors,

such as A.L. Kroeber and (especially) Edward Sapir, went on to

develop linguistics as an independent discipline in the USA. But thesingle most influential figure was Leonard Bloomfield, whose 1933

textbook Language effectively defined the field and set the agenda

for American linguists. Bloomfield’s successors, the Americanstructuralists (or post-Bloomfieldians), drew their inspiration from

Bloomfield, and they created a brand of linguistics which stressed

hands-on experience with real data and often dismissed the con-

temporary European tradition as mere ‘armchair theorizing’.

This is how things stood in the 1950s: a highly theoretical tradi-tion in Europe, a highly anti-theoretical tradition in the USA, and

something in between in Britain. But, in 1957, the young American

linguist Noam Chomsky published Syntactic Structures, a brief and

watered-down summary of several years of original research. In

that book, and in his succeeding publications, Chomsky made a

number of revolutionary proposals: he introduced the idea of a

generative grammar, developed a particular kind of generative

grammar called transformational grammar, rejected his Americanpredecessors’ emphasis on the description of data – in favour of a

highly theoretical approach based upon a search for universal

principles of language (later called universal grammar) – proposed

to turn linguistics firmly towards mentalism, and laid the founda-

tions for integrating the field into the as yet unnamed new dis-

cipline of cognitive science.

Chomsky’s ideas excited a whole generation of students; since

American universities were expanding rapidly in the early 1960s,these students quickly found jobs and began developing the field,

and within a few years Chomskyan linguistics had become the new

orthodoxy in the USA. Before long, Chomsky’s ideas had crossed

the Atlantic and established themselves also in many parts of

Europe.

Today Chomsky’s influence is undimmed, and Chomskyan lin-

guists form a large and maximally prominent cohort among the

community of linguists, to such an extent that outsiders often havethe impression that linguistics is Chomskyan linguistics, that lin-

guistics is by definition what the Chomskyans do. But this is ser-

iously misleading.

In fact, the majority of the world’s linguists would acknowledge

no more than the vaguest debt to Chomsky, if even that. Investi-

gators of historical linguistics, of sociolinguistics, of anthropological

LINGUISTICS

157

linguistics, of psycholinguistics and neurolinguistics, of language

acquisition, of dialectology, of semantics and pragmatics, of con-

versation analysis, discourse and texts, of computational linguistics,

and of a dozen other areas, all have their own agendas and prio-rities, and they are making progress – sometimes dramatic

progress – without paying any attention to Chomsky’s contribu-

tions. Indeed, it can reasonably be argued that the greatest advances

in our understanding of language in recent years have come from

the new field of sociolinguistics, in which pioneers like William

Labov, Peter Trudgill, and Jim and Lesley Milroy have transformed

our whole perception of what it means to speak a language.

Linguistics today is surely as lively a discipline as any on earth.In the past forty years or so we have probably learned more about

language than our ancestors managed in 2,000 years, and there is

no reason to believe that things are slowing down now.

See also: language

Further reading: Akmajian et al. 1995; Crystal 1997; Elgin 1983;Fromkin and Rodman 1998; Hudson 1984; O’Grady et al. 1996; Trask1995.

LITERACY

The ability to read and write effectively. Literacy is the ability to

read and write, and that sounds simple enough. But it isn’t.

Between the two extremes of a magisterial command of reading andwriting on the one hand and total illiteracy on the other, we find

any number of intermediate stages: literacy is a matter of degree.

One person may be able to read a popular tabloid newspaper but

unable to read a tax return or even the instructions on an aspirin

bottle. Another may be able to read a fair range of material but be

incapable of writing anything intelligible.

Attempts have been made since the 1940s at defining a level of

functional literacy – the minimum level of reading and writingrequired to function effectively in a world dominated by written

material – but this has proved exceedingly difficult to do. The

United Nations defines it relative to the culture in which the indi-

vidual lives, so the same functional literacy differs in absolute terms

between a nomadic African tribesman and an accountant living in

the print-rich West.

LITERACY

158

The term has also been extended to include notions such as cul-tural literacy, visual literacy, design literacy, computer literacy and

others.

Throughout history, the acquisition of literacy has been regardedas conferring high prestige (in many cultures, it has been the qua-

lification for movement from one social class to another, or the

qualification for electoral voting rights). Today, perhaps 75 per cent

of the world’s population may be described as literate – an enor-

mous advance, since it is not so many generations since a knowl-

edge of reading and writing was the exclusive preserve of tiny elites.

Mass literacy is largely a twentieth-century phenomenon.

An individual who lacks the reading and writing skills typical ofhis community is illiterate; an entire society which lacks a recog-

nized writing system is aliterate, or sometimes in a historical con-

text preliterate.

See also: oracy; standard language


LOAN WORD

A word copied into one language from another language. The

speakers of almost every language are in contact with the speakers

of other languages, and very often people take a liking to some of

the words used by their neighbours and take those words into their

own language. This process is called borrowing, and the words thatare taken over are loan words in the receiving language.

There are several motivations for borrowing a word. The simplest

is that the word is the name of something totally new to those who

borrow it. English, for example, has borrowed whisky from Scots

Gaelic, yogurt from Turkish, tomato from Nahuatl, sauna from

Finnish, ukulele from Hawaiian and kangaroo from the Guugu-

Yimidhirr language of Australia. The reason for this is that English-

speakers had never seen whisky or yogurt or tomatoes or saunas orukuleles or kangaroos before encountering these things overseas,

and so they simply took over the local names for them.

Another important motivation is prestige. At any given time in any

given place, some languages typically enjoy more prestige than others,

and speakers of less prestigious languages are often eager to show

off their command of a more prestigious language by introducing

LOAN WORD

159

some of its words into their own speech. For example, after the

Norman conquest of England, Norman French enjoyed far more

prestige than English, and English-speakers reacted by borrowing

huge numbers of Norman French words into English, such as pic-

ture, courage, army, treasure, language, female and even face, fool

and beef; in many cases these fashionable words simply displaced

their native English equivalents, which dropped out of use, as

happened with native here ‘army’ and andwlita ‘face’.

Loan words are often a good indication of social relations in

history. For example, it is likely that the large numbers of Scandi-

navian words (including, unusually, she and its, since pronouns are

rarely borrowed) which were borrowed by English indicate gen-erally good neighbourly relations between the natives and Viking

settlers in the east of Britain, instead of the pillaging raids depicted

in literature and film.

See also: language contact

Further reading: Blake 1996; Fennell 2000; Hock and Joseph 1996;Katamba 1994; Trask 1996.

LOCALIZATION

The concentration of certain mental faculties in particular areas of

the brain. We have long known that individual areas of the brain

are responsible for handling specific tasks. For example, not only

are there identifiable areas of the brain that process vision, there areeven areas that process specific aspects of vision, such as colour

vision and the perception of movement.

Language is no different: there are particular language areas in

the brain that process specific aspects of language use. In the mid-

nineteenth century, the French surgeon Paul Broca and the German

neurologist Carl Wernicke identified the two most important of

these: Broca’s area provides the grammatical structure for utter-

ances and sends instructions to the speech organs, while Wernicke’sarea handles comprehension of speech and access to vocabulary.

Both of these areas are usually located on the left side of the brain.

Broca and Wernicke reached their conclusions by examining

brain-damaged patients who exhibited specific disabilities of lan-

guage, and then by performing post-mortem inspections to deter-

mine which parts of the brain had been damaged. In recent years,

LOCALIZATION

160

however, the development of brain-scanning devices like PET(positron emission tomography), CT (computerized tomography) or

MRI (magnetic resonance imaging) scanners has allowed us to

monitor the brain of a healthy, conscious person who is performingsome linguistic task like speaking or reading; we can see directly

which parts of the brain are active during each of these tasks, and

as a result we now have a fairly detailed map of the language areas

in the brain.

See also: aphasia; language areas; neurolinguistics

Further reading: Akmajian et al. 1995; Ingram 2007; O’Grady et al.1996.

LOGONOMIC RULES

Very high-level rules governing discourse in a particular setting. In

a given setting, there are a number of high-level rules governing the

conduct of discourse: rules about who is allowed to speak andwhen, about who is allowed to interrupt or to change the subject,

about who is allowed to make jokes, about what counts as a joke,

about what is considered offensive, and so on. These are the logo-nomic rules governing that discourse, and they are so familiar to

participants (at least, to participants with experience of that type of

discourse) that they are virtually subconscious.

For example, the rules governing a seminar discussion in a uni-

versity are very different from those governing a performance by astand-up comedian in a night club. What is considered acceptable,

funny or offensive among a group of conservative farmers drinking

in a bar in South Africa or Montana will likely be very different

from the norms accepted by a lesbian or gay organization planning

a political campaign against discrimination.

Every normal speaker learns the rules appropriate to each type of

discourse in which he or she regularly takes part, and every normal

speaker can shift effortlessly from one set of rules to another aschanging settings require: this is part of that communicative compe-

tence which we acquire along with the acquisition of our purely

linguistic knowledge. And one of the most unpleasant experiences

we are likely to have is to be thrust into a setting of which we have

no experience and hence don’t know the rules or, still worse, in

which we disapprove of the rules.

LOGONOMIC RULES

161

An imperfect knowledge of the logonomic rules governing a set-

ting in which you find yourself will lead to your being regarded as,

at best, lacking in charm and social graces and, at worst, ignorant,

antisocial, stupid and offensive. Speaking out of turn, interruptingwhen you’re not entitled to, using a frostily formal tone in an

informal setting (or vice versa), attempting jokes which offend the

others – all these are failures to observe the logonomic rules and

evidence of your inadequate communicative competence.

See also: communicative competence; ethnography of communication

Further reading: Bonvillain 1993; Foley 1997; Harris 1993; Matthews1979; Newmeyer 1983.

MANNER OF ARTICULATION

The action of the speech organs involved in producing a particular

consonant. A consonant is produced by narrowing the vocal tract at

some point along its length. The particular speech organs chosen tomake the constriction represent the place of articulation, but, even at

a single place, it is usually possible to make several different kinds

of constriction. The type of constriction made in a particular

instance is the manner of articulation.There are several types of manner. In a plosive (like [b] or [k]), a

complete closure is made, blocking off the airflow, and the closure

is released suddenly. In an affricate (like [tS] or [ts]), a complete

closure is made and then released gradually, with friction noise. Ina fricative (like [f] or [z]), there is no complete closure, but air is

forced through a tiny opening, producing friction noise. These three

types are collectively called obstruents, because the airflow is

strongly obstructed.

The remaining types are collectively called sonorants. In a nasal(like [m] or [n]), a complete closure is made in the mouth, but the

velum is lowered, so that air flows out through the nose. In an

approximant (like [w] or most types of English /r/), the air isallowed to flow through a relatively large opening, and no friction

noise is produced. (At the phonetic level, such consonants are

strictly vowels, but they pattern in languages like consonants.) In a

flap (like the [�] of many languages of India), the tongue is ‘flipped’

rapidly from one place to another, briefly striking something else as

it moves. A tap (like Spanish [Q] in pero ‘but’) is similar except that

MANNER OFARTICULATION

162

the tongue finishes where it started. (Some books do not distinguish

between flaps and taps, but it is preferable to do so.) In a trill (likeSpanish [r] in perro ‘dog’), the air forced through a smallish open-

ing forces the tongue to vibrate. All these are examples of medianconsonants, in which all airflow is through the centre-line of the

mouth. However, it is also possible to block off the centre-line and

force the air to flow through one or both sides of the mouth; such a

consonant is lateral. We can produce a lateral affricate (like the [t ]

in Nahuatl, the Aztec language of Mexico), a lateral fricative (like

the [ ] of Welsh Llanelli), or a lateral approximant (commonly just

called a lateral) (like English [l]).

See also: consonant; phonetics; place of articulation

Further reading: Ashby 1995; Ball and Rahilly 1999; Ladefoged andMaddieson 1996; Laver 1994.

MARKEDNESS

The property which distinguishes less neutral linguistic forms from

competing ones which are more neutral. Though the concept is

older, the term markedness was introduced by the European lin-

guists of the Prague School in the 1920s, and it is now regarded as

linguistically central.

Markedness is a very broad notion applying at all levels of ana-

lysis. Generally speaking, a marked form is any linguistic form

which is less usual or less neutral than some other form – theunmarked form – from any of a number of points of view. (Note

that no linguistic item is absolutely neutral; neutrality is a relative

concept.) A marked form may be distinguished from an unmarked

one by the presence of additional linguistic material, by the pre-

sence of additional nuances of meaning, by greater rarity in a par-

ticular language or in languages generally, or in several other ways.

For example, voiceless vowels and voiceless laterals are marked

with respect to voiced ones, since the voiceless ones are far rarerthan the voiced ones in the world’s languages and since the voice-

less ones are generally found only in languages that also have

voiced ones, while most languages have only the voiced ones. The

affricate [pf], as in German Pflaume ‘plum’, is marked with respect

to both [p] and [f], since the last two are very frequent in languages,

while [pf] is exceedingly rare.

MARKEDNESS

163

English lioness is marked with respect to lion, since it contains

additional morphological material and since it is of less general

applicability. English brethren is marked with respect to brothers,

since the first (historically, simply the plural form of brother) isrestricted to certain special contexts, and bunny is marked with

respect to rabbit, since the first carries additional emotive meaning

absent from the second. A passive sentence like Janet was arrested

by the police is marked with respect to the active The police arrested

Janet, since the passive contains more material, has a more complex

structure, and is rarer than the active.

The several criteria for markedness may not always coincide. For

example, in some Pacific languages, passive sentences are far com-moner than active ones; hence the passive ones, although marked

from the point of view of grammar, are unmarked from the point of

view of discourse. Markedness values can also change over time: the

formal Latinate phrase prior to was once highly marked with respect

to native English before, but for very many speakers prior to has now

become the ordinary, and hence unmarked, form, as in prior to the war.

See also: phonology; sociolinguistics

Further reading: Hawkins 1984.

MARKER

In sociolinguistics, a marker is one of those features of a language

variety which are likely to be perceived by its speakers as having

social significance. For example, northern English speakers are

highly aware that their accents differ from southern English speak-

ers particularly in the pronunciation of the vowels in bus (N. /bUs/,S. /bvs/), grass (N. /græs/, S. /grA:s/) or boat (N. /bO:t/ or /bo:t/ or/bøt/, S. /b

eUt). Since they are usually above the level of conscious

awareness, markers are useful in sociolinguistic research where

speakers with different accents are observed accommodating (i.e.,

moving their accents towards each other).

In grammar, a marker is the affixed morpheme that indicates the

grammatical function of the word: such as the plural -s or past -ed

in English.

See also: accent; indicators; sociolinguistics; stereotypes

Further reading: Llamas et al. 2007; Stockwell 2002a.

MARKER

164

MEANING

The characteristic of a linguistic form which allows it to be used to

pick out some aspect of the non-linguistic world. The study of

meaning has a long history in a number of disciplines, notably

philosophy, but the linguistic study of meaning, semantics, largely

dates only from the late nineteenth century, and it has become gen-

erally regarded as a central part of linguistics only since the 1960s.

Linguists are chiefly interested in the meanings of linguistic forms

in everyday speech. More particularly, linguists are interested in theway some meanings relate to other meanings – that is, it is the

system of meanings which is seen as important, rather than the

meanings of individual items.

Like philosophers, linguists carefully distinguish different types

of meaning. The central and intrinsic meaning of a linguistic form is

its denotation or sense, while the fuzzy and sometimes variable asso-

ciations of that form constitute its connotations. The relation between

a form and the non-linguistic thing it picks out in a given context isits reference. An extended and non-literal meaning is a metaphor.

Debates in linguistics around meaningfulness have been con-

ducted, with formalist linguistics emphasizing the arbitrariness of

reference, cognitive linguists emphasizing its embodied and experi-

ential construal, psycholinguists stressing its dynamically con-

structed nature, and sociolinguists and discourse analysts pointing

to the performed and socially negotiated politics of the setting in

meaning construal.

See also: connotation; denotation; metaphor; reference; selection restriction;semantics; sense

Further reading: Allan 1986; Frawley 1992; Hofmann 1993; Hudson1995; Hurford and Heasley 1983; Leech 1974; Palmer 1976.

MEDIUM

Any one of several signalling systems within which a language maybe embedded. A primary medium for language is one which is not

derived from any other medium and which may be the medium in

which a child acquires its first language. The most familiar and

most widely used primary medium is speech, in which linguistic

information is encoded within sounds which are produced by the

vocal tract acting upon a stream of air. Speech has been used as a

MEDIUM

165

medium since our remote ancestors first evolved language, and it

was probably the first medium we ever used (though some scholars

query this). The other primary medium is sign language, in which

linguistic information is encoded within signs made chiefly with thehands, arms, shoulders, head and face. This is the medium used

today by the majority of deaf people, at least in places in which

they have a chance to learn it. Signing is of unknown antiquity; it

has become steadily more prominent since the eighteenth century,

but sign languages have very likely been constructed and used for a

very long time, whenever circumstances permitted. It is important

to realize that a true sign language is autonomous and in no way

derived from a spoken language.A secondary medium is one which is derived from a primary

medium. The most familiar secondary medium is writing, which

usually consists of an attempt at converting speech into permanent

marks. Writing is a recent invention in human affairs, being little

more than 5,000 years old, and until very recently knowledge of

writing (and reading) was confined to a tiny minority of the world’s

population. True sign languages can also be converted into writing,

though systems for doing this are still in their infancy and mostlyused only by specialists; most signers find it more convenient to

learn to read ordinary writing. Also secondary are systems like SignedEnglish (or Manually Coded English), in which a spoken language

(like English) is transferred into the medium of hand-gestures.

A tertiary medium is one derived from a secondary medium. The

slow and cumbersome system called finger-spelling is a tertiary

system, since it is derived from writing, with each letter of the

alphabet being rendered by a finger-sign. Morse code bleeps inelectronic form are tertiary. Early systems of shorthand were often

tertiary, being derived from writing, but modern shorthand systems

are all secondary, being derived directly from speech.

See also: literacy; sign language; speech; writing system

Further reading: Coulmas 1996; Crystal 1997; Trask 1995.

MENTAL SPACE

A conceptual representation of a semantic domain in cognitive lin-

guistics. In the mental space theory of French linguist Gilles Fau-

connier, the conceptual representation of the meaning of a sentence

MENTAL SPACE

166

is held in an idealized form. Further referential tokens in the dis-

course connect to the mental space and it is used to interpret the

utterance. The theory has been combined with conceptual metaphor

theory to explain metaphoric discourse in terms of several inputmental spaces producing a blended space. This new mental repre-

sentation is able to develop its own logic beyond the range of its

inputs. The theory of blending thus claims to account for the

transformative experience of metaphor.

See also: cognitive linguistics; metaphor; possible worlds theory; textworld theory

Further reading: Fauconnier 1994, 1997; Fauconnier and Turner 2002;Hickmann and Robert 2006.

MENTALISM

The doctrine that the mind can be invoked in scientific investigation

and even be made the object of study itself. In the late nineteenthcentury, psychology had become obscurantist and almost meta-

physical. The new doctrine of behaviourism attempted to sweep this

away by focusing only on what could be directly observed and pre-

ferably measured. But the behaviourists themselves often went so

far as not only to deny the possibility of appealing to unobservable

things like minds, but even to reject the very existence of minds.

More recently, the pendulum has swung the other way again, and

today most psychologists, philosophers and linguists are perfectlyhappy to invoke invisible things like minds and purposes and even

to make mind itself the object of study. This approach is called

mentalism, and it is now the dominant point of view in all three

disciplines, which themselves are now partly united within the new

discipline of cognitive science.

See also: behaviourism; cognitive linguistics

Further reading: Malmkjær 2004; Ungerer and Schmid 1996.

METALANGUAGE

A language used to talk about another language. Linguists,

philosophers and many others often need to talk about particular

METALANGUAGE

167

languages, or about languages in general. Naturally, the discussion

itself has to be couched in a language of some sort, and this fact

can quickly lead to hopeless confusion if we are not careful. We

must therefore distinguish carefully between the object language(the language which we are talking about) and the metalanguage(the language we are using to talk about the object language).

It is perfectly possible to use, say, English as a metalanguage in

order to talk about English as an object language, and indeed we

do this all the time, but it is precisely here that confusion can most

quickly arise: if we fail to distinguish between the English we are

talking about and the English we are using to talk about it, we can

easily become lost.Consider the following example. Using English as a metalan-

guage to talk about English, we may assert the following: a gram-

matical English sentence may not contain two consecutive instances

of the preposition of. This is true. But beginning students often

challenge this by pointing to examples like this very sentence: the

grammatical functions of of in English are numerous. Is the sen-

tence itself a counterexample?

No, it is not. It appears to be a counterexample only if we con-fuse the metalanguage with the object language. The statement

above is a statement in the metalanguage, and the first occurrence

of of in it is part of that statement. But the second occurrence of of

is different: this is merely a piece of the object language, one which

we happen to be talking about here. That is, the first of is doing

what of normally does, while the second of is doing nothing at all: it

is merely being talked about. If this is not obvious, observe what

happens when the object language is French: the grammaticalfunctions of de in French are numerous. Here it should be easy to

see that the French preposition de is doing nothing in this English

sentence except being talked about.

The two instances of of are an example of what philosophers

call the use-mention distinction: the first of is being used normally,

while the second is merely being mentioned (talked about): the

latter could appear in italics or in ‘scare quotes’ to signal this

graphically.At the very least, when we use English as a metalanguage to talk

about languages, we need to invoke a battery of technical terms and

concepts, just as a specialist studying physics, music or psychology

needs to invoke technical terms and concepts. In fact, we sometimes

go further, and invent a special artificial language to use as our

metalanguage. The various types of formal logic used by logicians,

METALANGUAGE

168

philosophers and linguists for describing English and other lan-

guages are just this: specially invented metalanguages.

See also: linguistics; philosophy of language

Further reading: Chapman 2000, 2006; Lyons 1995; Robson andStockwell 2005.

METAPHOR

The non-literal use of a linguistic form, designed to draw attentionto a perceived resemblance. The literary use of metaphors is ancient

and well studied, and the fields of rhetoric and literary criticism

have developed a formidable battery of Greek terms for naming

many different kinds of metaphor. But metaphors are in fact com-

monplace in ordinary speech and writing: we speak of the foot of a

mountain or the eye of a needle, we refer to Saussure as the father

of linguistics, and we speak of a failing business enterprise as a lame

duck. Like every language, English is stuffed with thousands of suchmetaphors, and most of them are so familiar that we no longer even

regard them as metaphorical in nature. Indeed, linguists have rea-

lized for some generations now that metaphors are a commonplace

way of extending the expressive resources of a language. And

essentially, of course, all referential linguistic tokens are metapho-

rical because of arbitrariness.

Since the early 1980s, however, those linguists developing the

new discipline of cognitive linguistics have been drawing attentionto the pervasive influence in languages of cognitive metaphors,large-scale metaphors which condition a broad range of expressions

and which appear to be related to the way we perceive the world.

An example is the underlying metaphor LIFE IS A JOURNEY, which

appears in any number of locutions: I’ll cross that bridge when I

come to it; She knows where she’s going; There were two paths open

to him; The baby has arrived; Their paths crossed; She has finally

arrived (succeeded); He is gone (dead); It’s been a long road; andmany others. A cognitive or conceptual metaphor is thus seen as a

mental mapping between two domains: a source domain of familiar

meanings and a target domain of the new meaning in focus. Many

of the tropes of classical rhetoric express in various stylistic forms

(such as synecdoche, litotes, hyperbole, metonymy, and stylistic

metaphor itself) aspects of an underlying cognitive metaphor;

METAPHOR

169

though it has been suggested that conceptual metonymy is a differ-

ent sort of mapping.

See also: cognitive linguistics; meaning

Further reading: Carter 1997; Cruse 1986; Giora 2003; Goatly 1997;Hiraga 2005; Lakoff and Johnson 1980; Kovecses 2005; Saeed 1997;Ungerer and Schmid 1996.

MINIMAL PAIR

A pair of words in a language which have different meanings but

which have identical forms except at one single point. The English

words pet and bet have different meanings, but they consist of

identical sequences of sounds in all positions except one: in this

case, initial position. Here the first word has [p], while the second

has [b]. Consequently, the difference in meaning must derive

entirely from the contrast between [p] and [b]; we must therefore

assign English [p] and [b] to different phonemes /p/ and /b/, and pet

and bet constitute a minimal pair for the phonemes /p/ and /b/.

Other minimal pairs for English /p/ and /b/ include planned and

bland, nipple and nibble, and rip and rib.

Finding a minimal pair for two sounds constitutes certain proof

that the two sounds in question must belong to different phonemes.

Sometimes this is fairly easy, as with English [s] and [S]: sun and

shun, puss and push, lass and lash. Sometimes it is not so easy, as

with English [S] and [Z], for which we can only find minimal pairsby invoking proper names (Aleutian versus allusion) or made-up

words (mesher versus measure). Larger minimal sets can be found.

For example, the set sum, sun, sung constitutes a minimal triplet for

the English phonemes /m/, /n/ and / /.

In sociolinguistics and in a slightly different sense, minimal

pairs of words are given to informants to test their accent fea-

tures, especially when the informants are acutely aware that their

speech is under examination: moon/moan; which/witch; blew/blue;law/lore.

See also: distribution; phoneme; phonotactics

Further reading: Carr 1993; Fromkin and Rodman 1998; Stockwell2002a.

MINIMAL PAIR

170

MINIMALIST PROGRAM(ME)

A version of generative grammar proposed by Noam Chomsky in

the mid-1990s. The emphasis in developing transformational andrepresentative rules should be on making them as economical as

possible. In other words, no rules should be redundant, and every

rule should be interpretable in the sense here of rendering a final

linguistic form. The deep structure and surface structure levels of

traditional generative grammar and in Government-and-Binding

Theory are removed to leave only the logical form and the phoneticform.

Transformational-generative grammar in the hands of Chomskycan be seen either as a unified and principled theory developing

with substantial shifts from the 1960s to the present, with Minim-alism as its latest version, or as a set of radically different revisions

that each amount to an abandonment of the previous system. Sev-

eral features remain consistent, however, to do with the emphasis

on syntax and phonology, the emphasis on generative rules, and the

delineation away from the sociolinguistic or performative context.

See also: Government-and-Binding Theory; transformational grammar

Further reading: Antony and Hornstein 2003; Chomsky 1995; Hendrick2003; Hornstein, Nunes and Grohmann 2005; Lasnik 2003; Lasnik etal. 2005; Radford 1997, 2004.

MINORITY LANGUAGE

A long-established language spoken as a mother tongue by people

in some part of a country in which the national or official language

is something else. We commonly tend to assume that everybody in,

say, France speaks French. In fact, virtually all adults in France do

speak French – but not always as their first language. In various

regions of the country, the first language of all or most local people

is Alsatian German, Dutch, Breton, Basque, Catalan, Occitan orCorsican. Each of these is the mother tongue in its region, and

people who learn one of these as their first language in early child-

hood may not even begin learning French until later in life, espe-

cially after beginning formal education.

Such languages are called minority languages, and minority lan-

guages are not confined to France. Alongside Spanish, Spain has

MINORITY LANGUAGE

171

Galician, Basque and Catalan; alongside German, Germany has

Frisian and Wendish (the second is a Slavic language); alongside

English, Britain has Welsh and Scots Gaelic; alongside English, the

USA has Navaho, Hopi, Lakota, and dozens of other indigenouslanguages, as well as other immigrant languages. Russia and China

each have over 100 minority languages, and countries with no sig-

nificant minority languages are in fact a rarity – though Iceland

and Portugal may be two cases.

In every case, a minority language has been spoken in its area for

centuries, sometimes even for millennia, but its speakers have

simply found themselves incorporated into a nation-state in which

the principal language is something else. In some cases, mostspeakers of a minority language may live and die without ever

acquiring an adequate command of the prestige language; in others,

all speakers normally become fluent in the prestige language, or at

least acquire an adequate working knowledge of it. Even in

Britain, Spain and France, monoglot speakers of Welsh, Basque

and Corsican were numerous only a few generations ago (and

Basque, at least, still has a handful of elderly monoglot speakers

today), but, because of dramatic advances in communications,transport and education, almost all European adults are today

fluent in the national language of the country they find themselves

living in.

This very fact, of course, places great pressure on minority

languages. In the not-too-distant past, speakers of minority lan-

guages like Welsh and Basque were often openly persecuted by

centralist governments, which were inclined to see the use of regio-

nal languages as unpatriotic, subversive, even dangerous. Todayopen persecution is less usual in Europe (though not always else-

where, and it has not disappeared entirely in Europe: for example,

Greece has recently been ferociously persecuting its Macedonian-

speaking minority). However, minority languages are nevertheless

often in grave danger of dying out. A good knowledge of the

national language is absolutely required for living in a modern

nation-state – to get an education, to find a job, to deal with tax

returns and innumerable other official documents, to travel roundthe country, to understand newspapers, television and films.

Further, knowledge of a major language like English or French

opens the door to opportunities, not just throughout the country,

but throughout the world. Consequently, speakers of minority

languages find themselves using the prestige language ever more

regularly, while the role of their mother tongue decreases

MINORITY LANGUAGE

172

correspondingly. In some cases, parents may actually strive to have

their children learn the prestige language, rather than their own

mother tongue, in the hope that these children will have a better life

as a result.Such pressures have already led to the disappearance of countless

minority languages, including Cornish and Manx in Britain and

hundreds of indigenous languages in North America and Australia.

Today these pressures are greater than ever, but in many cases

speakers of minority languages are waking up to the threat and

becoming increasingly militant in their demands for greater official

recognition and encouragement of their languages. In a few

cases, as with the Swedish-speakers of Finland and the Basque-speakers of Spain, a greater or lesser degree of protection has

already been achieved, but in most cases protection is limited or

non-existent, and it seems inevitable that hundreds or thousands

of minority languages will disappear in the next two or three

generations.

See also: language death

Further reading: Crystal 1997; Holmes 1992; Romaine 1994; Wardhaugh1987.

MIRROR NEURONS

It was discovered in the 1990s that the same brain neurons ‘fired’

both when a person performs an action and when that personobserves someone else performing the action. Most suggestively for

linguistics, these neurons are located in Broca’s area, a region

associated with language. The phenomenon of mirror neurons has

lent support to the gestural hypothesis of language evolution, which

proposes that language developed as a communication system out

of physical imitation in a community.

Furthermore, the mirror neuron effect occurs more strongly when

the observed action is tending towards an instrumental goal ratherthan an apparently unmotivated action. For example, more neurons

‘fire’ when the action involves picking up an apple and eating it

than when simply picking up an apple and moving it. By extension,

this seems to suggest that language evolved primarily as a goal-

driven mechanism (which would give support to schema theory),

and communication towards a purpose was a primary function of

MIRROR NEURONS

173

language origin. As with most theories of language evolution, con-

clusive evidence is hard to come by.

See also: cognitive linguistics; origin and evolution of language

Further reading: Goldstein 2002.

MODALITY

The grammatical category associated with the expression of obliga-

tion, permission, prohibition, necessity, possibility and ability. It is

by no means easy to separate modality from the more traditional

category of mood, which expresses degree or kind of reality. But forsome decades now linguists have preferred to apply the label mod-

ality to the linguistic expression of the six categories named above.

This term is particularly convenient in discussing a language, such

as English, which contains a specific set of modal auxiliaries for

expressing these concepts.

The English modals include can, could, may, might, will, would, shall,

should, must and ought (to), together with their negated forms. Famil-

iar examples like You can do it, You could do it, You may do it, Youshould do it, You must do it, You can’t do it and You mustn’t do it all

express aspects or degrees of one (or more) of the six named categories.

In practice, modality shades off imperceptibly into several other

categories: evidentiality (the amount and nature of the evidence

which you have for saying something), modalization (the probability

or regularity of an occurrence), modularity (the degree of commit-

ment or willingness involved on somebody’s part), hedging (redu-

cing your own commitment to what you are saying) and vaguelanguage. These extensions of modality have been particularly

investigated within Systemic Linguistics.

See also: mood

Further reading: Coates 1983; Frawley 2006; Hurford 1994; Palmer1974, 1979, 2001.

MODIFIER

A linguistic element in a sentence which is grammatically linked to

a second element and adds information about that second element.

MODALITY

174

A sentence is typically made up of smaller grammatical units called

phrases, and a phrase typically consists of a grammatically central

word, its head, accompanied by some modifiers, each of which pro-

vides some information about whatever is denoted by the head.Consider the sentence The little girl in the pond was shrieking

delightedly. Here the noun phrase the little girl in the pond has the

head girl (the whole phrase denotes some kind of girl), and girl has

two modifiers: little and in the pond. Each of these modifiers pro-

vides further information about the girl in question: among girls

generally, this particular one is both little and in the pond. The

word the, in contrast, is not a modifier but a specifier: it provides noinformation about the girl (it would make no sense to say that thisgirl is the compared to other girls), but rather it provides informa-

tion about how the entire noun phrase fits into the whole discourse

(in this case, it indicates that the girl in question has already been

mentioned earlier and is familiar to the listener).

Similarly, the verb phrase shrieking excitedly consists of the head

shrieking and the modifier excitedly – that is, of all the possible

ways of shrieking, this particular shrieking was done in an excited

manner.Modifiers can be quite large and complex, and they need not

occur immediately next to their heads. In the sentence The women

who had volunteered for the beauty contest climbed giggling onto the

stage, the head women is modified both by the relative clause who

had volunteered for the beauty contest and by the adjective gig-

gling, the second of which is separated from its head by the verb

climbed.

The relation between a modifier and its head is modification.

See also: head; phrase


MOOD

The grammatical category which expresses the degree or kind of

reality attached to an utterance. Mood is not well developed as a

grammatical category in English, and we mostly use words and

constructions to express mood distinctions. In this way, we can

assign varying degrees of certainty to an utterance: Susie smokes, I

hear that Susie smokes, It appears that Susie smokes, Surely Susie

MOOD

175

smokes, Susie probably smokes, Maybe Susie smokes, I wonder if

Susie smokes, It’s unlikely that Susie smokes. We can also assign

different degrees of reality within conditions: If Susie smokes, then

. . . (an open condition: maybe she smokes) is different from If Susie

smoked, then . . . (a counterfactual condition: she doesn’t smoke).

American English (but not British English) makes an overt mood

distinction between I insist that Susie smokes (I’m telling you: she

definitely does) and I insist that Susie smoke (I demand to see her

smoking – perhaps in a film role).

Questions may be regarded as expressing another kind of mood:

Does Susie smoke? And so can commands: Susie, smoke that cigar-

ette! (perhaps our tyrannical film director again, to a reluctantactress).

In some other languages, there are more or less elaborate dis-

tinctions of mood built into the grammar and frequently expressed

either by variation in verb-forms or by particles.

Mood shades off imperceptibly into modality, and no sharp line

can be drawn between them.

See also: modality

Further reading: Coates 1983; Hurford 1994; Palmer 2001.

MORPHEME

The smallest identifiable grammatical unit. We sometimes think of

words as the smallest units of grammar, but words in fact arenot the smallest grammatical units. For example, the word unhap-

piness is clearly built up from three smaller pieces: the prefix un-,

the stem happy and the suffix -ness. Each of these pieces is a mor-pheme of English, and not one of them can be broken down any

further in grammatical terms: all of them are minimal units (there

is no word hap to go with -y for example). Each of these mor-

phemes is used to build other English words. For example, the

prefix un- occurs also in unhappy, unclear, unwilling, uninterestingand unsatisfied; the stem happy occurs also in unhappy, happiness,

happier and happily, and also, of course, in the word happy itself;

and the suffix -ness occurs also in quickness, sadness, lewdness,

childishness and unpreparedness.

Morphemes are of different types. We say that happy is a lexicalmorpheme, meaning that it has dictionary meaning: we can provide

MORPHEME

176

a definition for it. But the -ly of happily is different: this is a

grammatical morpheme, which performs a strictly grammatical

function – in this case, that of turning an adjective into an adverb.

Quite independently, we can say that happy is also a free mor-pheme: it can stand alone to make a word, as it does in the word

happy. But the prefix un- and the suffixes -ness and -ly are boundmorphemes: they can never stand alone, but must always be

attached to at least one other morpheme within a word. In this

example, -y is not a morpheme within happy, though it is a mor-

pheme in shaky.

In the ideal case, a single morpheme has a single constant

form and a single constant meaning or function but, in practice,many morphemes vary in form, depending on where they occur.

For example, the morpheme sane has one form when it occurs in

the words sane and insane, but a different form when it occurs in

the word sanity (listen to the pronunciation, since the English

spelling is not very helpful here). Likewise, the negative prefix in-

exhibits several different forms in the words insincere, impossible

and illegal. We call these variant forms the allomorphs of the

morpheme.In the case above, it is a trivial matter to divide a word into the

morphemes of which it is built, but sometimes such division is not

so easy. Clearly the past-tense form loved (as in She loved him)

consists of two morphemes, the verb-stem love and a grammatical

morpheme which we can call Past, and it’s not too hard to draw a

line between them. But the past-tense form took must likewise

consist of two morphemes, the verb-stem take and the morpheme

Past, yet this time we can’t draw a neat line at all: the two mor-phemes are just wrapped up in a single bundle, and we have to

appeal to a more abstract level of representation to show that took

is really take plus Past.

The term morpheme was coined in the late nineteenth century by

the Polish linguist Jan Baudouin de Courtenay, but it was not

always used in the modern sense. That modern sense was estab-

lished by the American linguist Leonard Bloomfield in the 1930s,

and it was Bloomfield and his American successors who made thestudy of morphemes a central part of linguistics.

See also: affix; derivation; inflection; morphology; stem; word-formation

Further reading: Bauer 1988; Brown and Miller 1991; Jeffries 2006;Katamba 1994.

MORPHEME

177

MORPHOLOGY

Word structure, or the branch of linguistics which studies this.

Words typically have internal structure, and in particular they con-sist of smaller units called morphemes. For example, the verb-form

taking consists of two morphemes: the verb-stem take and the

grammatical ending -ing. Similarly, the noun textbook consists of

the morphemes text and book, and the the adverb slowly consists of

the morphemes slow and -ly.

Morphology is conventionally divided into two main areas. These

are inflection – the variation in form of a single word for gramma-

tical purposes, as with take, takes, took, taken, taking – and word-formation – the construction of new words from existing words, as

with textbook from text and book and with slowly from slow and -ly.

A particularly important type of word-formation is derivation, as in

slowly, rewrite and unhappiness.

See also: affix; derivation; inflection; morpheme; stem; word-formation

Further reading: Bauer 1988; Brown and Miller 1991; Katamba 1989,2006; Matthews 1991.

MOTHERESE

The distinctive style of speech used in addressing young children.

Adults looking after very young children, and above all mothersdoing this, typically talk to those children using a highly distinctive

type of language. They use special words like choo-choo, including

many diminutives like horsie and doggie; they confine themselves to

the simplest grammatical forms and constructions; they use exag-

gerated intonation patterns; they frequently repeat themselves; and

they often expand the child’s utterances into longer adult utterances

by responding, for example, to the child’s ‘Daddy sock’ with ‘Yes,

Daddy is putting his socks on’.This special type of speech is popularly known as ‘baby-talk’,

but this term is never used in linguistics; instead, it is variously

known as caregiver speech, caretaker speech or motherese. Lin-

guists are still debating the importance of care-giver speech in

allowing acquisition to occur. There is no question, however, that

many of the sociolinguistic norms for speakers are established by

MORPHOLOGY

178

the person who is most responsible for modelling their vernacular,

or mother-tongue.

See also: origin and evolution of language; vernacular

Further reading: Aitchison 1989.

MOVEMENT

Any of various processes or structures in which an element of a

sentence appears in a position other than its canonical position.Grammarians have long realized that some particular sentences

depart from the usual norms in the order of their elements, but it

was only in the 1950s, originally within the formalism of transfor-

mational grammar, that linguists began to speak, metaphorically but

conveniently, of the movement of elements out of their ordinary

position.

Movement phenomena are numerous and diverse. One familiar

example in English involves question-words (WH-words) like who

and what: we say Susie was talking to Mike, but we also say Who

was Susie talking to? and The guy who Susie was talking to is Mike,

in both of which the question-word who, which is logically the

object of the preposition to, fails to appear after to and instead

appears earlier in the sentence. This kind of movement is called

WH-movement or extraction.A second type of movement is extraposition, in which a complex

element is shifted to the end of its sentence. So, instead of That

Susie is drunk is obvious, we prefer to say It is obvious that Susie is

drunk, in which the that-clause has been extraposed. Similarly,

instead of A student who speaks Basque turned up this morning, we

can say A student turned up this morning who speaks Basque, in

which the who-clause is extraposed.

A third type is fronting, in which an element is moved to the

beginning of its sentence. So, instead of She inherited her brains from

her mother, we can say From her mother she inherited her brains, inwhich the phrase from her mother has been fronted. Likewise,

instead of I can’t cope with this we can say This I can’t cope with,

with fronting of this. When, as is often the case, an element is

fronted to make it an explicit topic, we speak of topicalization.Finally, there is a rather unusual type of movement called raising;

see that entry for information.

MOVEMENT

179

See also: raising; topic

Further reading: Kroeger 2004, van Valin 2001.

NAME

A linguistic form which serves to pick out a unique person, place or

thing. Grammatically speaking, a name is a noun phrase, but one

with the highly distinctive function of pointing at some individual

entity: Abraham Lincoln, Paris, the Golden Gate Bridge. (Tradition-

ally these have rather confusingly been called proper nouns.)Probably every person in every society receives a personal name

(or given name), typically shortly after birth, though in some socie-

ties an adult name is conferred later in life. In some societies, thereis a conventional list of personal names, one of which is selected; in

others, completely original names may be constructed and

bestowed. In the first case, it may be necessary to distinguish indi-

viduals bearing the same name by adding something to the name.

This can be done in several ways. A descriptive word may be used

to indicate something about the individual, such as personal char-

acteristics (size, complexion [Little, White]), profession (Miller,

Smith), location of dwelling (Atwater, Woods), place of origin(Bristol, Welsh), memorable incident; this is an epithet (or, if

humorous, a sobriquet). Or a patronymic may be added, identifying

the name-bearer’s father (Johnson, Prichard).

When such a second name ceases to be the property of a parti-

cular individual, and comes to be handed down to the name-bearer’s

descendants, it has become a surname (or family name). As can be

seen, many English surnames derive from epithets and patronymics.

Surnames in Europe are mostly of medieval origin, and even todaythey are not in use in Iceland and are little used in Turkish Cyprus.

Place names, or toponyms, are conferred everywhere upon every

kind of significant location: settlements, rivers, lakes, seas, valleys,

forests, fields, mountains and hills, roads and streets, bridges, city

gates, houses, places of worship, office buildings, sports stadiums

. . . the list is endless. Names are also conferred upon ships, trains,

planes, railway lines, festivals, holy days, books, newspapers . . . ,upon virtually everything that human beings consider important.

Name-giving practices differ considerably from society to society,

and anthropological linguists are often interested in studying these

practices.

NAME

180

Names exhibit some unusual linguistic properties, and their

study is a specialist subdiscipline within linguistics. The study of

names, and particularly of the origins and histories of names, is

onomastics.

See also: onomastics

Further reading: Crystal 1995, 1997; Dunkling 1995.

NARRATIVE

A text which tells a story. A narrative differs from most other types

of text in that it relates a connected series of events, either real or

fictional, in a more or less orderly manner. In addition to familiar

kinds of written narratives, such as history books and novels, there

are oral narratives, that is, stories told in conversation.

Narratives are of interest to linguists from various social,

anthropological and structural points of view. In the early 1970s,

the American linguist William Labov examined natural oral narra-tives and proposed a general six-part structure for them:

1 an abstract – indicating that a story is about to begin

2 an orientation – setting the scene and introducing the main

characters

3 the complicating action – the main events

4 a resolution – the outcome

5 an evaluation – explaining the point of the story, and6 a coda – signalling that the story is over.

It seems that this is the prototype form for narratives: deviations

from this form generate certain effects in the audience.

More recently, a number of linguists have devoted themselves to

the study of narratives of various types, and some of them have

attempted to develop story grammars, general structural outlines towhich particular types of narrative tend to conform. Naturally,practitioners of anthropological linguistics are interested in examin-

ing the rules for constructing narratives in different languages and

cultures. The field of narratology has developed as an interdiscipline

crossing linguistics with literary criticism and theory.

See also: genre; text

NARRATIVE

181

Further reading: Carter et al. 1997; Fabb 1997; Labov 1972; Palmer1996; Polanyi 1985; Propp 1968; Ryan 1991; Toolan 1988, 1994a, 1994b.

NATIONAL LANGUAGE

The single principal language of a country. By and large, we expect

citizens of France to speak French, even if their mother tongue is

something else. We expect to hear French spoken in the steeets, on

television and in films. We expect books and newspapers to be prin-

ted in French. We expect signs and advertisements to be in French,

and we expect to be able to conduct all our business in French,from buying a loaf of bread to making a will. In other words,

French is the national language of France – the single language that

is used by pretty much everyone for pretty much everything.

In the same way, English is the national language in Britain, in

Australia, in the USA and elsewhere, as German is in Germany and

Dutch in the Netherlands. But not every country has a national

language. Belgium is about equally divided between Dutch-speakers

(Flemish) and French-speakers (Walloon), and neither languagecan be used throughout the country. The same is true of Canada,

divided into English-speaking and French-speaking parts, and of

Switzerland, divided into regions speaking four different languages;

the same was true of Czechoslovakia and Yugoslavia before they

broke up, partly as a result of these very linguistic divisions. The

perception and encouragement of a single national language is

often a manifestation of a standard language ideology, where other

minority languages are suppressed, neglected or stigmatized.The position of a national language may be enshrined in law (as

in France) or not (as in the USA). But every country, whether it has

a national language or not, is obliged to recognize one or more

official languages in which official business may be conducted.

See also: minority language; official language; standard language

Further reading: Holmes 1992; Wardhaugh 1987, 2005.

NATURAL CLASS

A class of linguistic objects all of which behave in much the same

way in a language and which therefore often need to be referred to

NATIONAL LANGUAGE

182

in a description of the language. Every language contains a large

number of linguistic elements of various kinds: phonemes, mor-

phemes, words, and so on. Now, if every word in a language

behaved completely differently from every other word, and similarlyfor the other kinds of objects, then the language would be virtually

chaotic, and no organized description would be possible: we would

have nothing but a collection of miscellaneous observations.

But, of course, languages are not built like this at all. Instead,

almost all of these linguistic objects fall rather naturally into just a

few types, or classes, and all the items in one class exhibit very

similar behaviour. These classes are the natural classes of the

language.For example, almost all the words of English (or of any other

language) fall naturally into just a few classes: the parts of speech,

such as nouns, verbs and adjectives. Every noun behaves in much the

same way as every other noun, and so on for the other parts of

speech. Consequently, when we describe English, we can make a

number of important statements about the behaviour of the entire

class of nouns, and we do not have to concern ourselves with every

noun individually.The same goes for morphemes, phonemes and other classes of

objects. For example, the English phonemes /p t k/, as in pip, tit,

kick, form a natural class, since what is true of one of them is

nearly always true of the others as well, and so, by giving this class

a suitable name, the class of voiceless plosives, we can make all the

required statements about all of them at once.

There are two complications. First, it may be the case that, while

the members of a class behave identically in some respects, theybehave somewhat differently in other respects; this is the problem of

subcategorization, which is most prominent with parts of speech.

Second, our natural classes may sometimes overlap, so that a larger

natural class may contain within it several smaller natural classes

whose memberships are partly identical. This problem is most

familiar with phonemes, and it is the reason we prefer to work with

distinctive features.

One of the first goals in any description of a language is toidentify the natural classes that it contains and to provide suitable

labels for those classes, so that the description can then take

advantage of these classes whenever required. Failure to do this

leads inevitably to the endless unprincipled labelling of things, a

feature of traditional grammar widely regarded as one of its failings:

though the traditional grammarians certainly identified some natural

NATURAL CLASS

183

classes, they failed to recognize others that were just as important.

A large part of descriptive and theoretical linguistics in the twen-

tieth century has been devoted to the development of efficient and

principled ways of identifying, labelling and manipulating naturalclasses.

See also: distinctive feature; morpheme; part of speech; subcategorization

Further reading: Lakoff 1987; O’Grady et al. 1996.

NATURAL-LANGUAGE PROCESSING

The use of computer programs to process large quantities of lan-

guage data. Natural-language processing, or NLP, began to emerge

in the 1950s when high-speed computers first became available. The

original goal was machine translation (MT), constructing programs

that could translate a text from one language into another with a

minimum of human intervention. MT is still an active area today,

but most work in NLP is now concerned with a much broaderrange of applications.

Workers in NLP are concerned to devise computational techni-

ques for analysing substantial bodies of material in a natural lan-guage in order to obtain results comparable to those a human being

might obtain, but of course to do it very much faster and more

accurately. The central task is the construction of efficient and

robust parsers. A parser is a program which can take a sentence in a

natural language, work out its grammatical structure, and assign ameaning to it, so that the resulting meaning can then be manipu-

lated by other parts of the system.

Modern NLP systems are increasingly ambitious: they try to deal

not merely with single sentences but with sizable texts, often

including pragmatic and discourse factors. No system as yet comes

anywhere near the vast flexibility of language processing by

humans, but in particular domains NLP systems can be highly

successful. A familiar if simple example is the system used by travelagents for booking airline flights, but more elaborate systems are

also in regular use; for example, for extracting information from a

huge corpus of language data when preparing a dictionary, or in

internet search engines. NLP is now a central part of the enterprise

of constructing artificial intelligence, computer programs which can

mimic the behaviour of human beings in a range of areas.

NATURAL-LANGUAGE PROCESSING

184

(Note that the acronym NLP also stands for neuro-linguistic pro-gramming, a form of fairground quackery in which trauma victims,

addicts, depressives and others with serious mental disorders are

‘cured’ by repeating certain linguistic representations of their beha-viour: it is based on an extreme form of the Sapir-Whorf hypothesis.)

See also: computational linguistics

Further reading: Dale et al. 2000; Jurafsky and Martin 2000.

NEUROLINGUISTICS

The study of the connections between language and brain. The study

of the relation between language and brain was begun in the mid-

nineteenth century by the Frenchman Paul Broca and the German

Carl Wernicke. What they did was to study and characterize the apha-

sia (disturbed language) of people who had suffered brain damage, and

then, after the sufferers’ deaths, to conduct post-mortem examinations

in order to find out which areas of the brain had been damaged.In this way, they succeeded in identifying two specific areas of the

brain, today called Broca’s area and Wernicke’s area, each of which

is responsible for specific aspects of language use. These findings

confirmed the reality of the localization of language in the brain;

moreover, since these areas are nearly always located on the left side

of the brain, they also confirmed the lateralization of the brain.

In the mid-twentieth century, the American neurologist Norman

Geschwind elaborated the view of the brain as consisting of anumber of specialized components with connections between them,

and he also provided the basis of our modern classification of the

several language areas in the brain and of the types of aphasia

resulting from damage to each.

More recently, the introduction of sophisticated brain scanners

has allowed specialists to examine the activity in the brains of

healthy, conscious subjects who are performing specific linguistic

tasks like reading, speaking and listening. The new data have bothconfirmed and extended our understanding of the location and

functions of the several language areas.

See also: language areas; lateralization; localization

Further reading: Ingram 2007; O’Grady et al. 1996; Steinberg 1993.

NEUROLINGUISTICS

185

NEUTRALIZATION

The disappearance of a phonological contrast in a particular posi-

tion. It is clear that English has, among others, the two phonemes/p/ and /b/. These two distinguish a number of minimal pairs, such

as pie and buy, pike and bike, pray and bray, nipple and nibble, rip

and rib, and slap and slab. However, there is at least one position in

which they do not contrast at all: after an /s/ in the same syllable. In

English, we cannot possibly make a difference between two differ-

ent words such as spit and sbit, or spade and sbade.

In this position, therefore, we say that the contrast between /p/

and /b/ is neutralized: it no longer exists here and, in this position,/p/ and /b/ can no longer be used to distinguish pairs of words.

The concept of neutralization was introduced and developed

chiefly by the East European linguists of the Prague School in the

1930s, and especially by the Russian linguist Nikolai Trubetzkoy.

The existence of neutralization is a powerful reminder that the

phonology of a language is a matter of the behaviour and patterningof sounds, and not of their absolute phonetic value.

See also: minimal pair; phoneme; phonology

Further reading: Carr 1993; Clark and Yallop 1995; Gussmann 2002;Hawkins 1984; Katamba 1989; Lass 1984; Odden 2005.

NOMINALIZATION

Any grammatical unit which behaves like a noun or a noun phrase

but which is built up from something very different. The English

word arrive is a verb, as in She arrived at ten o’clock, but the word

arrival is a noun, as in Her sudden arrival surprised us. Clearly the

noun arrival is built up from the verb arrive, and so we say that

arrival is a nominalization of arrive.

This is an example of the simplest kind of nominalization, but

much more complex and elaborate types are possible. For example,in Tom and Sally Perkins study volcanoes, the sequence study volca-

noes is a verb phrase, but in Studying volcanoes is dangerous work,

this verb phrase has been nominalized into a noun phrase (it is the

subject of the sentence).

Yet another type of nominalization can be built up from Susie

smokes, which is a complete sentence. This entire sentence can be

NEUTRALIZATION

186

nominalized into a noun phrase, as in That Susie smokes surprises

me, in which the nominalization that Susie smokes is again the

subject of its sentence.

English allows adjectives to be nominalized only in limited cir-cumstances, as in The poor are always with us, in which the adjective

poor has been nominalized into a noun. But Spanish allows any

adjective at all to be nominalized, as in el rojo ‘the red one’, in

which the adjective rojo ‘red’ has been nominalized.

The use of nominalizations for various communicative purposes has

been particularly investigated within Systemic Linguistics, within which

nominalizations are treated as a kind of grammatical metaphor.

See also: critical discourse analysis; noun phrase

Further reading: Jackson 2002; Jeffries 2006.

NON-VERBAL COMMUNICATION

Any aspect of communication which does not involve words. Whenyou talk to somebody else, you stand or sit a certain distance away,

you adopt particular postures, you wear particular expressions, you

make particular gestures, and your choices often communicate

things that your words do not express directly: whether you are

interested or bored, whether you are nervous or confident, whether

you are attracted to the other person or not, and so on. All of these

are aspects of non-verbal communication, and many of them are

culture-bound, so that, when trying to speak a foreign language,you may inadvertently convey something you don’t intend.

The various non-verbal aspects of speaking are treated chiefly

under two rubrics: paralanguage for vocal but non-verbal behaviour

and the ethnography of communication for non-vocal behaviour.

With the easier use of portable video-recording equipment, the field

of kinesics has grown up to investigate the communicative effects of

body movement, especially accompanying verbal communication.

(Note carefully that, in language studies, the term verbal means‘expressed in words, either spoken or written’; the everyday sense of

this term to mean ‘spoken, oral’ is never used in linguistics.)

See also: ethnography of communication; paralanguage

Further reading: Poyatos 1993, 2002.

NON-VERBAL COMMUNICATION

187

NOTATIONAL CONVENTION

Any recognized shorthand for stating a linguistic rule briefly. Lin-

guistics makes heavy use of notational conventions, some of themuniversal, others confined to particular theoretical frameworks. In

phonology, the rule ‘The consonant /n/ is lost between vowels’ may

be written as follows: n ! ø/V____V. In syntax, the rule VP ! V

NP (PP) means ‘a verb phrase may consist of a verb plus a follow-

ing noun phrase, with an optional prepositional phrase at the end’.

Mastery of these conventions is essential for work in descriptive

and theoretical linguistics; they not only provide a systematic and

standard disciplinary language, they also allow general rules to bestated precisely without interference from the surface words them-

selves.

See also: formal grammars; International Phonetic Alphabet; metalanguage;rule

Further reading: Aarts and McMahon 2006.

NOUN

The part of speech which includes words like girl, tree and happi-

ness. Traditional grammarians often tried to define a noun as ‘the

name of a person, place or thing’, but this doesn’t work. Clearly, for

example, red is the name of a colour, and so, by this definition, it

should be a noun – and yet it is most usually an adjective, as inSusie is wearing her red skirt.

Like any part of speech, nouns can be adequately defined only in

terms of their grammatical behaviour. In English, an obvious

grammatical characteristic of nouns is that most of them can

appear in two different grammatical forms, called singular and

plural. Most English nouns form their plural by adding -s, as in girl/

girls and tree/trees, but some have irregular plurals deriving from

historically fossilized forms, as in child/children, goose/geese, sheep/sheep or from other languages by borrowing: radius/radii. However,

not all nouns do this: some have only a singular form (like wheat,

furniture and the anglicized form of the Italian plural spaghetti),

while others have only a plural form (like scissors, police and pants –

though some retailers will engage in a bit of back-formation to try

to sell you their new pant).

NOTATIONAL CONVENTION

188

A better way of identifying nouns is to use a suitable gramma-

tical frame. Consider the two frames The – was nice and The – were

nice. If you can put a single word into one of these blanks to make

a good sentence, then that word must be a noun, because thegrammar of English allows nouns, and only nouns, to appear in

these positions. The first frame accepts singular forms of nouns, like

girl, spaghetti and furniture, while the second accepts plural forms,

like trees, pants and police. (Of course, there is no guarantee that

the result will be sensible: The torture was nice doesn’t sound very

normal, but it’s clearly grammatical, and so torture is a noun.)

See also: noun phrase; part of speech

Further reading: Hurford 1994; Jackson 2002; Jeffries 2006; Mahlberg2005.

NOUN PHRASE

A syntactic unit which can act as a subject or an object. Considerthe following unremarkable sentence: A small party of Spanish

adventurers managed to capture the Aztec capital. A traditional view

of sentence structure holds that the subject and the direct object of

the sentence are the nouns party and capital, respectively. But this is

not strictly correct. The true subject or object are the phrases a

small party of Spanish adventurers and the Aztec capital. These are

noun phrases (or NPs for short), and noun phrases are the only

things that can act as subjects or objects in English sentences. Anoun phrase is a syntactic unit – a constituent – and a noun phrase

may be identified in two different ways. First, it must occupy one of

only a few possible slots in a sentence structure. Second, it must

have one of only a few possible types of internal structure.

With only a couple of exceptions, an English noun phrase is

always built up around a single noun, and that noun is the head of

the noun phrase, the item which is chiefly responsible for the nature

of the NP. In my examples, party and capital are the heads of thetwo NPs. The first NP denotes a particular party, and the second a

particular capital, and the other words in the NP serve only to

provide further identification.

The most obvious exception is a noun phrase consisting of a

pronoun. In the sentence They managed to capture it, the pronouns

they and it make up complete NPs all by themselves, one serving as

NOUN PHRASE

189

the subject, the other as the object. This is what a pronoun typically

does: it makes up a noun phrase all by itself, and a pronominal NP

is the most familiar kind of NP (in English) which is not built

around a head noun (the facts are different in some other languages).A noun phrase, then, is so called because it is (usually) built

around a noun and because it constitutes a complete syntactic unit,

a phrase. A noun phrase is one kind of syntactic category.

A simple but convenient way of testing whether some phrase is a

noun phrase is to try inserting it into a suitable frame to see if the

result is a grammatical sentence. Here’s an example of a frame:

– am/is/are nice. Any syntactic unit which can fit into the blank

successfully must be a noun phrase, because it will have to be thesubject of the sentence. (The reason we need several possible verb-

forms is that an English verb shows agreement with its subject.) So,

all of the following are NPs (or at least they can be NPs): she, spa-

ghetti, this little book, most of the other students in my class, the

woman I was talking to, Rome and Paris.

The reason for the qualification – ‘or at least they can be NPs’ –

here is the following. In Spaghetti is nice, spaghetti forms an NP all

by itself, but in This spaghetti is nice, the subject NP is this spa-

ghetti, and this time spaghetti is not a noun phrase, but only a part

of a noun phrase.

As happens with any kind of phrase, a noun phrase may contain

within itself a smaller noun phrase. In my example a small party of

Spanish adventurers, this big NP contains within it the smaller NP

Spanish adventurers (try this in the frame). This NP is doing

another typical NP job: it is serving as the object of the preposition

of, and the resulting prepositional phrase has been incorporatedinto the bigger NP.

Among the several possible structures for an NP, by far the most

frequent is a combination of a determiner with a certain other syn-

tactic category; this other category is called an N-bar (or sometimes

a nominal group). Examples (in each case, the first bracketed item is

the determiner and the second is an N-bar): [that] [girl]; [two] [little

puppies]; [the] [woman in the blue skirt]; [a] [book (which) I’m read-

ing]. Note that girl in the first example is both a noun and an N-bar, and note that the other examples definitely have the structure

shown; for example, the structure *[the woman] [in the blue skirt] is

quite wrong for the third example, as shown by the asterisk.

Introductory textbooks often decline to recognize N-bars when

drawing structures for NPs, in the hope of making life simpler for

the reader: however, N-bars are essential for all serious work on

NOUN PHRASE

190

English sentence structure. Note also that, within the Government-

and-Binding Theory, noun phrases have recently been re-named

determiner phrases (DPs), for theory-internal reasons.

See also: determiner; noun; phrase; pronoun

Further reading: Greenbaum and Quirk 1990; Jackson 2002; Jeffries2006.

NUMBER

The grammatical category which relates most directly to the number

of entities. For human beings, the number of people or objects

under discussion is often of some importance, and our languages

typically provide us with a rich vocabulary for making limitless

distinctions along this dimension: none, one, two, three, twenty-

seven, one-half, 0.42, about a hundred, some, few, many, no more than

four, a handful, and so on. So far these distinctions have nothing to

do with grammar. But it is perfectly possible for a language to buildsome of these distinctions into its grammar – not all of them, of

course – and a language which does so has the grammatical cate-

gory of number.English, like most European languages, has the category of

number, but English has only a very simple contrast between a sin-gular (representing one entity) and a plural (representing two or

more entities). Hence nouns denoting things which can be counted

typically have two forms: dog/dogs, child/children, cactus/cacti.Nouns denoting things that cannot be counted typically have only

one form, most often one which is treated grammatically as a sin-

gular: water, happiness, disgrace, wheat (but note that oats has only

a plural form). A few words, though, are idiosyncratic and excep-

tional, such as furniture, which has only a singular form, and pants,

which has only a plural, even though both nouns appear to denote

things we could reasonably count.

Observe that the English number system makes no explicit pro-vision for zero or for fractions, and here we must arbitrarily choose

either a singular or a plural form: we say nobody is ready, with a

singular form, but we agonize in other cases: should we say none of

the students is ready or none of the students are ready?

Some other languages make more elaborate distinctions of

number than does English. In Arabic, for example, a noun typically

NUMBER

191

has three forms: malikun ‘king’, malikani ‘two kings’, malikuna

‘three or more kings’; the second form is called the dual. (The

English comparative-superlative system has something of this in

distinguishing the latter or the worse from the last or the worst in alist). The Pacific language Larike has four forms for pronouns:

mane ‘he’ or ‘she’, matua ‘they two’, matidu ‘they three’, mati ‘they

(four or more)’; the third form is the trial. The East African lan-

guage Tigre has a different system: faras ‘horse’, ?afras ‘a few

horses’, ?afresam ‘horses’; the second form is the paucal.In still other languages, the category of number is absent alto-

gether, at least for nouns. Mandarin Chinese is one such. The Chi-

nese word shu- ‘book’, like every noun, has only the one unvaryingform, and there is nothing in Chinese corresponding to the English

book/books distinction. When distinctions of number are important,

they must be expressed in words: ‘one book’, ‘two book’, ‘many

book’. Some English nouns, especially in dialectal variation or fos-

silized forms, are similarly invariant: two foot (as a measurement of

length), ten-year-old, threepenny.

Even though we give the name number to the grammatical cate-

gory, it is important to realize that grammatical number is some-thing quite different from real-world number: it represents no more

than an (always imperfect) attempt to map certain real-world dis-

tinctions into the grammar.

See also: grammatical category; person; tense


NUMBER OF LANGUAGES

A count of the total number of distinct languages spoken as mother

tongues. It is by no means a simple matter to count the world’s living

languages. To start with, of course, we suffer from incomplete

information. Until very recently, large parts of the planet were

poorly investigated, and we simply had little or no informationabout the languages spoken there. This was particularly true of

New Guinea, Australia, south-east Asia, many parts of Africa and,

above all, the Amazon basin. Today, thanks to the dogged efforts of

hundreds of linguists, we are much better off than we were a gen-

eration ago, but, even so, a previously unknown language still turns

up occasionally: one was discovered in Brazil in 1995.

NUMBER OF LANGUAGES

192

But there is a much bigger problem. Our familiar picture of

Europe, in which a single language is spoken over hundreds of

miles and shows rather sharp boundaries with neighbouring and

quite distinct languages, is a recent development in human affairsand is still today not typical of most of the globe (and even parts of

Europe are not like this). Far more typically, what we find is a dia-lect continuum. That is, as we travel across the terrain, the local

speech just changes gradually, and we appear to be looking at

nothing more than regional dialects of a single language. But, over

a sufficient distance, we find that the differences in the local vari-

eties are so great that speakers of these distant varieties cannot

understand one another at all. That is, all speakers can talk easilyto their near neighbours, and with more difficulty to more distant

neighbours, but not at all to those still farther away.

The question then arises: how many languages are we looking at?

And where should we draw the lines between them? There is no

principled answer to such questions, and, in practice, linguists just

have to do the best they can. Naturally, different linguists looking

at the same part of the world do not always come to the same

conclusions; furthermore, when better information becomes avail-able, all of the earlier decisions may have to be revised, and lin-

guists may not agree about what revisions to make, either. Hence,

we will get different assessments of the number of languages spoken

in the area depending on whose work we are reading.

What makes Europe different is the rise, in the past few centuries,

of nation-states with central governments and well-defined bound-

aries. These political boundaries have largely imposed upon the

dialect continua of Europe a substantial degree of order: in prac-tice, today, what language you speak depends largely on no more

than what side of a boundary you live on.

Consider Dutch and German. For centuries, a single Germanic

dialect continuum has covered much of northern and central

Europe. The local varieties spoken on both sides of what is now the

Dutch–German frontier were not significantly different, but local

varieties spoken farther apart were often very different indeed: even

today, two speakers from, say, Berlin and Heidelberg cannot under-stand each other at all if they use their own local varieties. The

greatest differences, in fact, have always been north–south: the

Netherlands, Belgium and northern Germany on the one hand

versus southern Germany, Switzerland and Austria on the other.

But the political frontiers have been drawn, and today two

speakers born a kilometre apart, but finding the Dutch–German

NUMBER OF LANGUAGES

193

frontier separating them, consider that they speak two different

languages: Dutch in one case, German in the other. And mass

education, of course, reinforces this perception: one speaker learns

standard Dutch in school, the other standard German, the samestandard German being learned by other speakers hundreds of

kilometres away whose mother tongue is incomprehensibly

different. Hence the ‘official’ decision is that we are looking at just

two languages, and the enormous regional variation is quietly

ignored.

The local varieties spoken in Belgium, Luxembourg, Alsace,

Switzerland and Austria are, again, all incomprehensibly different

from all the other varieties and from one another. And again thedecisions are political. The Swiss and the Austrians have decided

that they too speak German. The Luxembourgers, after much

vacillation, are now seemingly concluding that they do not, and

they are making efforts to establish their Letzebuergesch as a dis-

tinct language. The Belgians have long insisted that they spoke a

distinct language called Flemish, but they have now changed their

minds and agreed that they speak Dutch – but note that the local

speech of western Belgium is incomprehensible to speakers inAntwerp and in Amsterdam. The Alsatians, with long-standing

political grievances against the Germans, also consider that they

speak a separate language – though, if Germany had succeeded in

its repeated attempts to annex the territory, the outcome might be

very different.

Similar political decisions are important elsewhere. Varieties of

Finnish, Romanian and Persian are spoken in sizable areas of the

former Soviet Union, but, for political purposes, the Soviet autho-rities always insisted that the varieties on their territory were

entirely separate languages, called Karelian, Moldavian and Tadjik

respectively. If, as most linguists do, we follow the Soviet line, then

we get three additional languages that we would not have got if

Soviet policy had been different.

So how many languages are there? Most estimates have ranged

between 5,000 and 6,000, or occasionally as low as 4,000 or even

3,000. The most authoritative source we have at present, the Eth-

nologue volume (Grimes 2000), currently recognizes just over 6,500

mother tongues, though a more recent study based in Wales and

associated with UNESCO has reported the astounding total of just

over 10,000 languages. However, because of accelerating language

death, this total is declining rapidly, and more than half of these

languages may be gone within a century.

NUMBER OF LANGUAGES

194

Globalization of travel and communication is to blame: even

though this apparently offers more opportunities for language con-

tact (which could lead to pidginization and thence creoles), in prac-

tice the dominance of English as a lingua franca seems to be havingthe opposite effect.

See also: language death; national language; standard language

Further reading: Crystal 1997, 2003; Grimes 2000; Krauss 1992.

OBSERVER’S PARADOX

The problem for data collection in linguistic fieldwork when the

presence of the researcher affects the form of language being used

by the informants. People being recorded are often likely to be self-

conscious and this can have an effect on their linguistic perfor-

mance. Interactants in an extended setting usually accommodatetowards each other’s language patterns, so the researcher’s language

variety is likely to influence the speech under investigation. Even anartificial test situation in itself can skew results without the creator

of the test being present in the room.

Various methods in sociolinguistics have been developed to mini-

mize the effects of the observer’s paradox. Misleading informants

into thinking that their knowledge is being tested rather than the

way they express themselves, or surreptitious recordings have been

used, though these are increasingly ethically questionable. Reducing

the formality of the setting by using relaxed seating, casual dress,informal interviews or conversations has been successful. Matching

the researcher with the informants in terms of age, gender or eth-

nicity has also been used. Using protocols that engage the infor-

mants in emotionally involving responses (the ‘danger-of-death’

question: tell me about a time when you feared for your life),

spending an extended period living amongst the informants (parti-cipant observation), recording events which are already public and

being recorded (such as media broadcasts, court proceedings orpolitical debates) – all these have been used to minimize the potential

interference of the research process itself.

The question of ethics in linguistic fieldwork research is raised by

the issue. Some have argued that the desire to minimize or even

eradicate the observer’s paradox represents a desire to write out

the messiness of human informants in the research process, and

OBSERVER’S PARADOX

195

sociolinguists ought to be more humane and sensitive to linguistic

performance. Others argue that the phenomenon means that only

extended qualitative research can be valid, rather than the sort of

empirical and artificial setting which tends to be the hallmark ofquantitative work.

See also: sociolinguistics

Further reading: Labov 1994; Llamas et al. 2007; Milroy 1987b.

OFFICIAL LANGUAGE

A language which can be used for conducting official business in a

particular country. In a modern nation-state, every one of us is

constantly obliged to engage in some kind of official business with

the authorities. We have to get driving licences and passports, fill in

income-tax returns, obtain birth certificates for our children, fill in

forms both when we obtain jobs and when we become unemployed,

make out wills, get our children into school, buy and sell houses,and so on. Some of us have to participate in local or national gov-

ernment, or to appear in court. For these purposes, every govern-

ment specifies one or more languages in which such business may

legally be conducted, and a language singled out in this way is an

official language within the territory of that government. But which

language or languages should be official in a given country?

In some cases the choice is easy. Since Swedish is the first lan-

guage of practically all native-born Swedes, Swedish is the onlyreasonable choice for the official language. In other cases, the

choice is more difficult. Belgium is about equally divided between

Dutch-speakers and French-speakers, and the Belgian government

has been obliged to draw up complicated laws about which lan-

guages can be used officially in which parts of the country. Spain

has long recognized nothing but Spanish – the majority language –

as official, but constant objections from the millions of Spanish

citizens whose first language is Catalan, Galician or Basque have nowpersuaded Madrid to set up autonomous regions with their own

regional governments, several of which recognize the local language

as co-official with Spanish. Nigeria is inhabited by speakers of

dozens of different languages, and the government has reacted by

choosing English – the language of the former colonial power – as the

official language, to avoid antagonizing any part of the population.

OFFICIAL LANGUAGE

196

Attempts at making Hindi the sole official language in India have

encountered fierce resistance from the hundreds of millions of

Indians who speak other languages. After independence from Britain

in 1948, English was supposed to be phased out, but it has proven auseful ‘neutral’ language as it does not privilege any native ethnic

group; so in practice English remains the official language there too.

See also: minority language; national language

Further reading: Holmes 1992; Llamas et al. 2007; Stockwell 2002a;Wardhaugh 2005.

ONOMASTICS

The study of names. Onomastics is a branch of philology, and its

pursuit requires the same painstaking scrutiny of historical docu-

ments as any other branch of philology, especially since names have

a habit of changing more dramatically and more irregularly than

ordinary words. For example, until the documents are consulted, it

is far from obvious that the name of the English village of Bridg-water originally meant ‘bridge [at a place owned by a man named]

Walter’, that the name of the American river the Picketwire con-

tinues an earlier French Purgatoire ‘Purgatory’, or that the district

of London called Pimlico derives its name in a complex fashion

from the North Carolina river now called the Tar-Pamlico.

Onomasticians study both anthroponyms (personal names, espe-

cially surnames, but also given names) and toponyms (place names);

toponyms include habitation names (names of settlements), hydro-nyms (names of bodies of water), oronyms (mountain names), and

the names of valleys, fields, roads, streets, houses, forests, and any

other features that can be named.

See also: name; philology

Further reading: Crystal 1995; Trask 1996.

OPEN-ENDEDNESS

The ability to use language to say new things, without limit. A non-

human species typically has no more than a handful of messages,

or calls, available to it – perhaps one meaning ‘This is my terri-

tory’, another meaning ‘danger in the air’, and so on. Every time a

OPEN-ENDEDNESS

197

non-human creature opens its mouth, it can do no more than to

choose one message from this short list, and that’s it. A monkey

may be able to say ‘Snake!’, if that message is available in the system,

but that same monkey cannot possibly produce an unprecedented‘Look out – two hunters with rifles!’ or, still less, on spotting its first

Land Rover, ‘Hey, everybody – what do you suppose that is?’

Human languages are utterly different. We have not the slightest

difficulty in producing and comprehending totally new utterances

that we have never used or heard before, and indeed we do this

almost every moment: most of the utterances you produce and hear

every day have very likely never before been produced by anybody.

Consider a few examples: A large tear rolled down the little pink

dragon’s nose; Peanut butter is a poor substitute for putty; Lux-

embourg has declared war on New Zealand; Shakespeare wrote his

plays in Swahili, and they were translated into English by his African

bodyguards. You have no difficulty in understanding these – even if

you don’t believe all of them.

But even your most routine utterances may never before have

been uttered by anybody: I ran into Susie’s ex-husband at the match

last night; Does anybody know what language the word ‘shampoo’

comes from?; Aunt Bea has sent us some photos of her grand-

daughter’s christening.

This limitless ability to produce and understand totally new utter-

ances is called open-endedness, and it should be perfectly clear to you

that, without it, our languages and indeed our lives would be unrec-

ognizably different from what they are. Perhaps no other feature of

language so dramatically illustrates the vast, unbridgeable gulf separat-

ing human language from the signalling systems of all other creatures.The importance of open-endedness has been realized by linguists

for decades; the term was coined by the American linguist Charles

Hockett in 1960, though others have sometimes preferred the labels

productivity or creativity.

See also: design features; displacement; stimulus-freedom

Further reading: Carter 2004; Maybin and Swann 2006.

OPTIMALITY THEORY

A development of generative grammar formulated by Alan Prince and

Paul Smolensky in the 1990s. The theory holds that all languages

OPTIMALITY THEORY

198

have a set of constraints which produce the basic phonological and

grammatical patterns of that particular language. In many cases, an

actual utterance violates one or more of these constraints, so a

sense of well-formedness applies to that utterance which violatesthe least number or least important constraints. Constraints can be

classified in two types: faithfulness and markedness. The faithful-

ness principle constrains a word to match the underlying morpho-

logical form (such as plural tram +-s in trams). But words like buses

or dogs do not follow this constraint (the first falls foul of the con-

straint that prevents the pronunciation of two consecutive /s/

sounds and the second places a /z/ instead of an /s/). These two

examples, though, follow markedness constraints, and in these casesthe particular markedness ‘scores’ higher than the faithfulness

constraint, so the alternate forms are allowed. Differences between

languages, then, are a matter of the relative importance given to

particular constraints, and a description of these constitutes a

description of the language.

See also: minimalist programme; phonology

Further reading: Archangeli and Langendoen 1997; Blutner and Zeevat2003; Dekkers et al. 1999; Kager 1999; Kuhn 2003; McCarthy 2001,2004; McMahon 2000.

ORACY

Skill in speaking and listening. We have long had the term literacyto denote skill in reading and writing, but recently a number of

educationalists and academics have been drawing attention to the

importance of skill in speaking and listening. To this notion the

somewhat unfortunate label oracy has been applied, though a few

people, disliking this term, use oralcy or orality instead.

Sociolinguists and anthropological linguists have discovered that

oral skills can be deeply important: oral skills are highly valued in

any number of societies and may confer great prestige. In Americaninner-city ghettos, gang members can acquire status by their ability

to hurl biting insults; in small African and Asian communities, the

most effective speakers are likely to be the headmen and chiefs.

In English, the evaluation of oral skills is complicated by the

powerful tendency to judge speakers entirely in terms of their use of

standard English: so, very often, a piece of pompous, turgid and

ORACY

199

disorganized standard English is rated more highly than a sharp

and effective piece of speaking delivered in a conspicuously non-

standard variety.

See also: literacy

Further reading: Halliday 1989; Ong 1982; Tarleton 1988.

ORIGIN AND EVOLUTION OF LANGUAGE

The series of steps by which human language came into existence.Very little is known about how human language came into exis-

tence, though there is no shortage of speculation by specialists in a

dozen different disciplines.

The members of most non-human species have some way of

communicating with their fellows, and mammals in particular typi-

cally use a combination of calls (vocal signals) with postures, ges-

tures and expressions. Proponents of the continuity hypothesis see

language as deriving directly from such systems by simple elabora-tion. But most linguists and many others see these signals as more

akin to such non-linguistic activities as sobbing, laughing and

screaming, and they prefer to invoke discontinuity hypotheses, bywhich language has an entirely different origin, one not detectable

in living non-human species.

Specialists differ in the date they assign to the rise of language.

The most popular view sees language as arising with our own spe-

cies, Homo sapiens, a little over 100,000 years ago. But someanthropologists believe they can detect evidence for language areas

in the brains of our hominid ancestors of 1 million or 2 million

years ago, while, on the other hand, some archaeologists argue that

full-blown language can only have arisen around 40,000–50,000

years ago, a time when they see evidence for a spectacular flowering

of art, culture and material goods.

There is no question that language has given humans a huge

evolutionary advantage. Much of the physical evidence is in theform of feedback circularity: standing up on two legs raised our

sensory organs for more input and freed our hands up; having fin-

gers in front of our eyes allowed us to iconically represent numbers

(up to 10) which developed our memories and abstract reasoning

and grew our brains; increasingly subtle dexterity promoted brain

growth; balancing a head on shoulders allowed the head to be

ORIGIN AND EVOLUTION OF LANGUAGE

200

bigger; the head grew in size because the jaw, forehead and vocal

tract enlarged; having children with large heads while two-legged

mothers had smaller pelvises meant that babies had to be born

before their ‘full’ term development was complete; so babies arehelpless in comparison with other mammals, and language gives us

a means of passing on childcare knowledge and increases the indi-

vidual’s chances of surviving infanthood; and so on.

While specialists in some other disciplines often like to portray

language as something not very different from what vervet monkeys

do, and hence as requiring a minimum of explanation, almost all

linguists are satisfied that human language is in fact dramatically,

utterly different from everything else we can see, that language,more than anything else, is what makes us human, and that the

origin of language is therefore a major problem which we are not

close to solving. Most linguists further believe that our language

faculty is genetic in nature, that our remote ancestors simply

evolved it, and hence that we are born to use language in the way

that birds are born to fly.

See also: animal communication; bioprogram hypothesis; genetic hypothesisof language; innateness hypothesis; protolanguage hypothesis

Further reading: Aitchison 1996; Crystal 1997; Foley 1997; Malmkjær2004.

ORTHOGRAPHY

A standardized system for writing a particular language. A givenorthography consists of a particular version of a particular writing

system (in the case of English, a version of the Roman alphabet

employing both small letters and capital letters), a standardized

system of spelling, often a system of word-division (in English, by

white spaces), and (in almost all cases today) a particular system of

punctuation. Often there are additional devices, such as (in English)

the use of the arabic numerals (1, 2, 3, . . .) for writing numbers, the

specialized use of certain symbols (such as +, = , &, %, $, @), thepresence of certain abbreviations (such as Dr, St, etc., e.g.), and the

use of distinctive typefaces like italic and boldface; these devices

may or may not be regarded as strictly a part of the orthography,

but more commonly are not.

There are numerous orthographic systems in use around the

world. Various alphabetic systems such as the Roman, Arabic and

ORTHOGRAPHY

201

Cyrillic use graphemes to represent (roughly) phonemes. Logo-graphic systems use pictures or symbols to represent entire words

as well as the sounds occurring in those words: as in Egyptian

hieroglyphics or modern Chinese. Syllabary systems use a singlesymbol to represent a syllable (typically a consonant plus a vowel),

such as Mayan, Cherokee and much Japanese. The last of these is

more complex, since Japanese writers can use logograms borrowed

from China called kanji, as well as two syllabic systems katakana

and hiragana, and also some adapted Roman symbols called

romaji.

The orthography of English has developed gradually over the

centuries. The English spelling system was largely fossilized byprinting around 600 years ago, so much modern English ortho-

graphy actually represents medieval pronunciation (through, knight,

folk): this is quite useful as it means modern spelling does not pri-

vilege any single accent, and English writing can be used to write

any pronunciation. However, written texts from King Alfred’s time

to the present nevertheless show numerous orthographic changes.

Letters have been added to or removed from the alphabet: early

printing presses did not have the Anglo-Saxon <æ> so a <Y> wasused with a point ground out of the block at the top, leading to

spellings such as ‘Ye old shop’, which was always pronounced as

‘the’ and never /ji/. The forms of certain letters (notably <s> as <Q>or <f>) have often varied, and the spellings of individual words

have changed (sunne to sun, coude to could by mistaken analogy

with would and should). Punctuation practices have varied enor-

mously; the use of capital letters has varied; roman numerals have

mostly given way to arabic numerals; individual abbreviations havecome and gone; and so on. Consequently, if you want to read texts

in Old English or Middle English, you must master not only the

language but the orthography as well.

See also: punctuation; spelling; writing system

Further reading: Crystal 1995, 1997; Joshi and Aaron 2006; Malmkjær2004; Robinson 1995; Sampson 1985.

PARADIGM

A complete set of the inflected forms of a single word, especially as

an example. In foreign-language teaching of a traditional kind, it is

PARADIGM

202

commonplace to present a full set of the various grammatical forms

assumed by a single word, with the understanding that this set

constitutes a model for the behaviour of a whole set of words

belonging to a single grammatical class. Such a set of forms is asingle paradigm. So, for example, a textbook of Latin might provide

the paradigm for the noun amicus ‘friend’, as follows; the various

functions of the named case-forms must be learned separately, of

course:

(This is the American order; British tradition puts the accusative

straight after the vocative.) The point of this is that all the other

Latin nouns belonging to this class, the class of ‘masculine second-

declension nouns’ or ‘masculine o-stems’, behave in exactly the

same way, so that, once you know the forms of amicus, you auto-matically know how to make all the forms of all these other

nouns (apart from any which are irregular and must be learned

separately).

Note that paradigms are not confined to nouns: verbs, adjectives

and other parts of speech may also exhibit paradigms, sometimes

very elaborate ones.

In modern language-teaching, paradigms are less prominent than

formerly, but they are still routinely provided for reference. Indescriptive work in linguistics, however, paradigms can still be an

illuminating and economical way of presenting the morphological

facts of a language in which words vary their forms substantially

for grammatical purposes.

The branch of morphology that deals with paradigms is called

morphosyntax to signal that there is a strong relationship with

syntax.

See also: inflection; morphology; paradigmatic relation

Further reading: Blake 2001; Culler 1986.

Singular Plural

Nominative amicus amiciVocative amice amiciGenitive amici amicorumDative amico amicisAccusative amicum amicosAblative amico amicis

PARADIGM

203

PARADIGMATIC RELATION

The relation between a set of linguistic items which, in some sense,

constitute choices, so that only one of them may be present at a

time in a given position. The notion of a paradigmatic relation was

introduced by the Swiss linguist Ferdinand de Saussure in the early

twentieth century as a generalization of the traditional concept of a

paradigm. If you look at the several forms of the Latin noun amicus

listed under paradigm, you will realize that, in any given position in

a Latin sentence in which the word amicus is present, one, and onlyone, of the several forms will occur, the choice being determined by

the grammatical context.

What Saussure did was to point out that the relation among the

several forms in a traditional paradigm is essentially the same as

the relation among other sets of linguistic elements. So, for exam-

ple, the English determiners, like a(n), some, the, this/these and that/

those, all stand in a paradigmatic relation, because, in a single noun

phrase, there will be one and only one of them at the beginning (asa rule: there are exceptions): a book, some books, the book, this

book, those books, and so on. It is for this reason that all these

words are assigned to a single part of speech. In large measure, all

the members of any part of speech are assigned to that part of

speech because they are related in the same way: they all occur in

the same positions, but only one of them can occur at a time.

The concept of a paradigmatic relation is closely related to that

of a system, a set of competing choices and the rules for choosingamong them, and the term paradigmatic relation contrasts most

obviously with syntagmatic relation.

Note that, in psycholinguistics, the term paradigmatic relation is

used in a somewhat different sense, to denote the mental associa-

tions between words which form part of a set of mutually exclusive

items, as when a subject given black responds with white.

See also: paradigm; syntagmatic relation

Further reading: Culler 1986.

PARALANGUAGE

The non-linguistic aspects of speaking. When we speak, of course,

we communicate a good deal of purely linguistic information to our

listeners. In addition, however, we make use of strictly non-linguistic

PARADIGMATIC RELATION

204

variables like pitch, loudness, tempo, timbre and voice quality. Our

use of these things conveys information about our mood and atti-

tude: about whether we are angry, amused, nervous, excited, impa-

tient, tired or whatever. These aspects of speaking are collectivelycalled paralanguage or, informally, tone of voice.

Such paralinguistic features as high pitch, falsetto, creaky voice, a

‘gravelly’ voice, breathy or whispery voice, nasalization, and loud or

soft speech are variously used in many languages to indicate

respect, submission, mockery, boredom, romantic or sexual feelings,

impatience and many other things; the details differ greatly from

language to language.

Note that the term paralanguage is sometimes used more nar-rowly, to include only voice quality, and sometimes more broadly,

to include most or even all aspects of non-verbal communication.

The sense given here is recommended, since suitable terms are

already available for the narrower and broader senses.

The equivalent of paralanguage can also be observed in sign

language: signers may produce signs rapidly or slowly, more delib-

erately or more casually, using large movements or small ones.

See also: ethnography of communication; non-verbal communication

Further reading: Poyatos 1993, 2002.

PAROLE

The real utterances produced by real people in real situations. The

term parole was introduced by the Swiss linguist Ferdinand deSaussure in the early twentieth century; in Saussure’s treatment, this

term contrasts specifically with langue, the abstract system of a

language. Saussure’s parole represents essentially the same thing as

Chomsky’s performance, though his langue is significantly different

from Chomsky’s competence.

See also: langue; performance

Further reading: Culler 1986.

PARSING

Traditionally, parsing was an exercise in the language classroom for

labelling the parts of speech in a sentence (noun, verb, adjective, or

PARSING

205

nominative, accusative, dative, and so on). In modern linguistics, it is

associated with the attempt towards natural-language processing in

computational linguistics. A parser program depends on a rule-

driven model of grammar to be able to tag an input sentence withits constituent structure. Phrase-structure grammars are popular as

the basis for computer parsers.The result turns the natural language

string into a hiererchically and systemically organized set of data,

which is then available for interrogation by, for example, concordanceprograms or further software that aims to ‘understand’ and ‘interpret’

the sequence in order to generate a ‘reply’. The numerous ‘scare

quotes’ here are to signal the fact that, although this aspect of

artificial intelligence research has progressed to the point at whichvoice-operated computers are commonplace, there is still a long

way to go before a computer program passes the Turing test. (Thisis a simple test devised by the British mathemetician and crypto-

grapher Alan Turing in the 1950s aimed to replace the question of

whether a machine could ever think by the more measurable ques-

tion of whether a machine can appear to have natural language. If a

person conversed with two entities in another room, one a human

and the other a computer, and could not tell the difference, then themachine would have passed the Turing test. Several people have

noted that this is not quite the same as ‘intelligence’, but it remains an

as yet-unreached landmark in natural-language processing research.)

See also: case; computational linguistics; phrase-structure grammar

Further reading: Heil 1998; Nivre 2006; Turing 1950.

PART OF SPEECH

Any one of the grammatically characterized classes into which the

words of a language are grouped. Every language contains many

thousands of words. If all these words behaved differently for

grammatical purposes, the language would be unmanageable; but

they don’t. Instead, they are grouped into a small number of clas-ses, variously called parts of speech or word classes or lexical cate-gories; the words in each class behave in much the same way, while

words in different classes behave differently. Not all languages have

the same classes: some classes, such as noun and verb, appear to be

universal, while others, such as adjective and preposition, are found

in some languages but not in others.

PART OF SPEECH

206

English has over a dozen parts of speech; the precise number

varies according to the analysis, since some linguists prefer to draw

finer distinctions than others. The meaning of a word is an unreli-

able guide to its part of speech. Membership is determined bygrammatical criteria, and there are at least three types of criteria to

which we can appeal.

The first criterion is distribution, the positions in which a word

can occur. For example, consider the following ‘frames’: This – is

good; These – are good. If we try to put words into the blanks to

produce good sentences, we find that nouns, and only nouns, will

work, because English grammar allows only nouns in this position.

Hence spaghetti, dogs and arrangement are nouns, while happy, fromand deliver are not. (The reason we need two frames is that English

nouns occur in two forms: singular and plural; see number.)

The second criterion is inflection. In English, a typical noun has

exactly two grammatical forms: singular and plural: dog/dogs, box/

boxes, child/children, radius/radii. (A few, though, have only one or

the other, such as happiness, furniture, oats and police.) An adjective

may have three forms: big/bigger/biggest. A verb usually has four or

more: love/loves/loved/loving; take/takes/took/taken/taking. A prepositionhas only a single form: to, under, without.

The third criterion is derivation. A noun may take the suffix -ful

or -ous to form an adjective: joy/joyful, power/powerful, glory/glor-

ious, mountain/mountainous. An adjective may take the suffix -ness

or -ity to form a noun: great/greatness, topical/topicality. A verb

may take the prefix re- or un- to form a different verb: write/rewrite,

do/undo. In each case, only a member of the appropriate class can

take that prefix or suffix successfully.These tests are not infallible, and they must be used with care.

For example, you might think that the frame This is a – dress would

pick out adjectives, and it’s true that adjectives will go in here: red,

short, pretty. But nouns will equally go into this position: cotton,

maternity, cocktail. A better frame here would be This is a very –

dress, which does indeed pick out only adjectives.

A feature of English and of some other languages is that a word

can belong to two or more different classes without changing itsform. For example, brown is a noun in a nice shade of brown, an

adjective in a brown skirt, and a verb in Please brown the meat.

Likewise, straight is a noun in Schumacher accelerated down the

straight, an adjective in a straight line, and an adverb in She hit the

ball straight. There are many of these, but note that a word can only

belong to one part of speech at a time. Some other languages do

PART OF SPEECH

207

not tolerate this and require each word to belong only to one part

of speech.

As a rule, the words in a single class do not all show absolutely

identical behaviour; instead, they are further divided into severalsubclasses, often overlapping, which show somewhat different

behaviour. This is subcategorization.

Some word classes are large and can readily accept new mem-

bers: these are called open classes. Others are small and accept new

members only with difficulty: these are closed classes. In English,

noun, verb and adjective are open classes, while pronoun and pre-

position are closed classes. Languages may differ here: in some

languages, the class of adjectives is small and closed.

See also: adjective; adverb; conjunction; determiner; noun; preposition;pronoun; verb

Further reading: Crystal 1995; Givon 1993; Hurford 1994.

PERCEPTUAL STRATEGY

Any of several rough principles which listeners may use in interpreting

utterances. A perceptual strategy is essentially a kind of principled

guess about how the words we are hearing fit into a syntactic structure.

In the past several decades, practitioners of psycholinguistics have

proposed a number of such strategies, and the reality of these stra-

tegies is supported by varying amounts of experimental evidence.

One of these is the principle of late closure, which says ‘if possi-ble, put the next word into the phrase you are currently processing’.

By this principle, if you hear Susie decided gradually to get rid of her

teddy-bears, you will associate gradually with decided, and not with

to get rid of her teddy-bears.

Another proposed strategy is the canonical sentoid strategy, bywhich the first string of words that could possibly be a sentence is

assumed to be a sentence. It is this strategy which makes garden-path sentences so difficult to interpret: when you hear The horse

shot from the stable fell down, you naturally take The horse shot

from the stable as a complete sentence and are left floundering by

the continuation, even though the whole utterance has a perfectly

straightforward interpretation.

Perceptual strategies have attracted a great deal of attention, and

they have enjoyed some success in accounting for the comprehension

PERCEPTUAL STRATEGY

208

of short utterances. With long and complex utterances, however, it

becomes very difficult to identify any useful strategies, and some

workers are now questioning the utility of perceptual strategies as a

tool for investigation.

See also: language processing; performance; psycholinguistics

Further reading: Aitchison 1989.

PERFORMANCE

Real utterances produced by real people. When we speak, our

utterances are frequently disturbed in various ways. We make slips

of the tongue; we forget things; we pause to consider our words; we

start an utterance and then break it off; we may even lose track of

what we are saying altogether. Likewise, when we listen, we may fail

to catch something, or we may mishear it or misunderstand it.

Linguists have long realized that these disturbances to our linguistic

behaviour largely result from non-linguistic causes, and hence thatthey should not be treated on the same footing as the linguistic

behaviour itself. Consequently, we make a fundamental distinction

between our (somewhat idealized) capacity to use language, called

langue by Saussure and competence by Chomsky, and our actual

linguistic behaviour, called parole by Saussure and performance by

Chomsky.

Linguists with a theoretical orientation are usually interested in

competence, and so they disregard what they see as the irrelevant‘noise’ of speech errors. But there are other linguists, especially

those interested in language processing, who are deeply interested in

speech errors, and these may be fairly said to be engaged in the

study of performance.

See also: competence; language processing; langue

Further reading: Steinberg 1993.

PERFORMATIVE

An utterance which is itself an act of doing something. A typical

utterance, such as I’m going to a film, cannot readily be regarded as

PERFORMATIVE

209

doing something: saying this does not constitute going to a film.

But some utterances are different. Saying I promise to buy you a

teddy-bear, all by itself, constitutes making a promise to buy a

teddy-bear, and nothing further is required to complete (as opposedto fulfil) the promise. An utterance of this sort is called an (explicit)performative utterance, and a verb which can be so used, in this case

promise, is a performative verb.Further examples of performative utterances are I now pronounce

you husband and wife and I hereby name this ship HMS Pooty. In

these cases, clearly, the utterances will have no effect unless a

number of obvious conditions are met (the speakers are a minister

of religion or a governmental representative on one hand, and aperson with some prestige status on the other; the settings are

a wedding and a shipyard, and so on). These are the felicityconditions for that utterance. If the felicity conditions are not satis-

fied, then the resulting utterance is not really false or wrong: it is

merely infelicitous, and it has no effect (or at least not the intended

effect).

The existence of explicit performative utterances was first pointed

out by the British philosopher J.L. Austin in the 1960s. Austin’swork has been continued by several linguists and philosophers, and

it has led to the development of the theory of speech acts.

See also: speech act

Further reading: Austin 1962; Saeed 1997; Thomas 1995; Yule 1996.

PERSON

The grammatical category relating to differing roles in speech. The

(personal) pronouns of English make a three-way distinction: the

pronoun I means ‘the speaker’, you means ‘the addressee’, and he,

she and it all mean ‘somebody or something else’. We say that these

forms distinguish persons: first person for I, second person for you,

and third person for the others. The category represented is alsocalled person, and this three-way person contrast appears to be

universal in languages.

Observe that the third-person forms express further distinctions

of animacy and sex, but these other distinctions have nothing to do

with person. Finnish han and Turkish o mean ‘he’ or ‘she’ indiffer-

ently; these languages lack the English sex-distinction. Other

PERSON

210

languages, though, make sex-distinctions in their word for ‘you’

and sometimes also for ‘I’.

The plural forms of the personal pronouns are more compli-

cated. The first-person plural we does not normally mean ‘thespeakers’; instead, it means ‘the speaker and one or more others

associated with the speaker’, and something similar is often true

when you is used as a plural. (Among the world’s languages, modern

English is unusual in not formally distinguishing a singular you

from a plural you, though of course in older and modern dialectal

forms there are distinctions between thou and you, or you and y’all.)

Some North American and Pacific languages distinguish two sets

of third-person forms, one (the proximate) serving to indicate theindividual who is currently the centre of attention and the other

(the obviative) indicating all individuals who are not currently the

centre of attention. The obviative has sometimes been labelled the

‘fourth person’, but this term is not obviously appropriate.

See also: grammatical category; number; tense


PHILOLOGY

The branch of historical linguistics concerned with the histories of

individual words and names. Historical linguists exhibit a range of

interests. Some are chiefly interested in determining which lan-

guages are connected in genetic relationships and in working out thestructures of particular language families; others are concerned with

identifying the particular changes which have affected individual

languages; still others are looking for general principles of language

change. But some are mainly interested in identifying the origins of

particular words or names and in tracing the histories of these

items through time. We give the name philology to this pursuit, and

we divide philology into etymology – the study of the origins of

ordinary words – and onomastics – the study of the origins ofnames.

Philology has been around a long time and, until not so long

ago, this term was often applied to historical linguistics generally –

though no longer. Philological work is exacting, and it requires a

mastery of detail perhaps greater than in any other branch of

linguistics. Few other areas of the subject offer so many ways for

PHILOLOGY

211

the unwary practitioner to go astray: overlooking a single crucial

datum may lead to catastrophe, and that datum may lie buried in

an obscure publication or manuscript in a library hundreds of miles

away. Nevertheless, philologists have achieved prodigious success intheir undertakings, and the great etymological dictionaries of Eng-

lish, Spanish, Latin and other languages are among the treasures of

linguistics – though all of them contain errors which remain to be

corrected by later scholars.

See also: etymology; historical linguistics; language change; onomastics


PHILOSOPHY OF LANGUAGE

The branch of philosophy which studies the properties of human

languages. There are many aspects of language which are of equal

interest to linguists and to philosophers, particularly (though not

exclusively) in the domain of semantics (the study of meaning).Philosophers of language are often interested in such questions as

how a piece of language can refer to the real or conceptual world,

how the truth or falsehood of a statement can be determined, how

the meaning of an utterance depends upon its context, and what the

relation is between language and mind.

Though philosophers have pondered problems of language since

ancient times, the subject was particularly stressed in the late nine-

teenth century by the German philosopher Gottlob Frege, who isconsequently often regarded as the father of the discipline. Among

the most prominent philosophers of language in the twentieth cen-

tury are the Britons Bertrand Russell, Peter Strawson, Donald

Davidson, J.L. Austin, Paul Grice and John Searle, the Pole Alfred

Tarski, and the Americans Willard van Orman Quine, Richard

Montague, Jerrold Katz, Jerry Fodor and Saul Kripke, but there

are many others who also deserve mention.

(Special note: there is a school of philosophy called linguisticphilosophy, so named because its proponents maintain that many

philosophical problems arise from insufficient attention to lan-

guage; this has nothing to do with the philosophy of language and

should not be confused with it.)

See also: semantics

PHILOSOPHY OF LANGUAGE

212

Further reading: Chapman 2000, 2006; Martin 1987; Robson andStockwell 2005; Stainton 1996.

PHONATION TYPE

Any one of the several different ways the vocal folds may behave in

producing speech. The vocal folds (or vocal cords) are two move-

able masses of tissue in the larynx between which air from the lungs

must flow during speech. The vocal folds can behave in a number of

different ways, sometimes with different parts of them doing dif-

ferent things. Most familiarly, they can be brought close enoughtogether to vibrate along their entire length, producing voicing, or

they can be moved far apart, preventing vibration and producing

voicelessness, or, more precisely, breath. They can also be pressed

tightly together, blocking all air flow and producing a different type

of voicelessness.

We can also close the vocal cords except for an opening at the

back through which air flows noisily; this is whisper. Or we can

close them apart from an opening at the front which vibrates veryslowly; this is creak. Or we can stretch them tightly along their

entire length so that they vibrate very rapidly; this is falsetto. Var-ious combinations of these are possible, such as whispery voice,creaky voice, whispery falsetto and whispery creaky voice (‘whisky

voice’).

Not all of these possibilities are used for linguistic purposes, but

several of them are. English contrasts only voicing and voiceless-

ness, but many other languages have a third possibility, most oftencreaky voice or whispery voice.

See also: airstream mechanism; International Phonetic Alphabet; voicing

Further reading: Ball and Rahilly 1999; Ladefoged 1993, 1971; Ladefogedand Maddieson 1996.

PHONEME

One of the basic sound units of a language. Every spoken language, or

more precisely every distinguishable accent of a language, possesses a

smallish set of basic abstract sound units, both consonants and

vowels, and every word in that language must consist of a permitted

PHONEME

213

sequence of those basic sound units, which are called phonemes. Thenumber of phonemes in particular languages varies considerably,

from a known minimum of ten in the Brazilian language Piraha to

a known maximum of 141 in the African language !Xu. The aver-age number seems to be about twenty-five to thirty.

The several accents of English vary noticeably in their set of

phonemes, from as many as forty-five in some accents of England

to as few as thirty-six in some accents of North America. These

differences lie mostly in the vowels, with most varieties having

exactly twenty-four consonant phonemes, though a few varieties

differ even here.

Phoneme symbols are always enclosed within slashes, and thesymbol chosen for each phoneme is usually a phonetic symbol

intended to suggest how that phoneme is most typically realized

phonetically. So, for example, the phoneme that occurs at the

beginning of thin is commonly realized as a voiceless dental fricative

[h], and so the phoneme is represented as /h/. Likewise, the entire

word thin is represented as /hIn/, with three phoneme symbols for

the three phonemes present in it, while think is /hI k/, and this is

/DIs/.A crucial point is that a single phoneme need not always get the

same phonetic realization. English /p/, for example, is phonetically

an aspirated plosive [ph] in pin, but an unaspirated plosive [p] in

spin. Phonetically, then, we have [phIn] and [spIn], but phonemically

we have merely /phIn/ and /spIn/: the phonetic difference is pre-

dictable, and it ‘does not count’ in English. There is only one pho-

neme /p/ here, and English-speakers typically do not even notice the

phonetic difference; we say that [ph] and [p] are allophones of thephoneme /p/. In contrast, Mandarin Chinese has pairs of words like

[pha-] ‘crouch’ and [pa-] ‘eight’, and so [ph] and [p] clearly belong to

two different phonemes, /ph/ and /p/, and these words are phone-

mically /pha-/ and /pa-/ (the diacritic is a tone mark).

English has pairs of words like [den] den and [Den] then, and so

[d] and [D] clearly belong to two different phonemes, /d/ and /D/,and these two words are phonemically /den/ and /Den/. In Spanish,

however, things are different. Spanish has a single phoneme /d/,which phonetically is [d] after pause but [D] between vowels. Thus,

dedo ‘finger’ is phonemic /dedo/ but phonetic [deDo]; dama ‘lady’ is

phonemic /dama/ and phonetic [dama]; but la dama ‘the lady’ is

phonemic /la dama/ but phonetic [la Dama].

The phoneme concept was worked out only slowly in the nine-

teenth century; the two Polish linguists Jan Baudouin de Courtenay

PHONEME

214

and Miko�aj Kruszewski are usually credited with being the first to

understand it fully. The concept was carried to the west, where it

was championed by Daniel Jones in England and by Edward Sapir

in the USA; by the 1930s it was almost universally understood andused in linguistics.

In the 1960s, American linguists, led by Noam Chomsky and

Morris Halle, began to develop and defend a much more abstract

conception of the phoneme than had formerly been normal, within

the new framework called generative phonology; their new concep-

tion was dubbed the systematic phoneme, as opposed to the tradi-

tional autonomous (or classical) phoneme. Moreover, the phoneme,

which had previously been regarded as an indivisible minimal unit,was now decomposed into a matrix of smaller units, the distinctive

features, which were increasingly regarded as the true fundamental

units of phonology. Since the 1980s, this trend has continued to the

point at which many phonologists now work exclusively with fea-

tures and ignore phonemes altogether – though an understanding

of phonemes is still considered essential in most introductory

linguistics courses.

See also: distribution; minimal pair; neutralization; phonology; phono-tactics; transcription

Further reading: Gussmann 2002; Lass 1984; Odden 2005; Sommerstein1977.

PHONETICS

The study of speech sounds. Phoneticians (practitioners of pho-

netics) investigate such topics as the anatomical, physiological and

neurological basis of speech (this is physiological phonetics), the

actions of the speech organs in producing speech sounds (articu-latory phonetics), the acoustic nature of the sound waves which

transmit speech (acoustic phonetics), and the manner in which the

ears and brain interpret speech (auditory and perceptual phonetics).Phoneticians have long used various mechanical devices in their

investigations; today they more commonly use a battery of electro-

nic instruments backed up by computers, and most phonetics today

is therefore instrumental phonetics. Those phoneticians who prefer

to work entirely by ear, without instruments, are said to be doing

impressionistic phonetics.

PHONETICS

215

A modern phonetician is expected to be at home in a laboratory

full of instruments, but is nevertheless also expected to undergo a

good deal of ear training, acquiring the ability to recognize and

characterize speech sounds entirely by ear. Also necessary is amastery of the International Phonetic Alphabet, the standard system

for transcribing speech sounds.

Phoneticians may choose to investigate either the total range of

speech sounds which can be produced by human beings (this is

general phonetics or anthropophonics), or the manner in which

speech sounds are used in real human languages (this is linguisticphonetics, and it overlaps with phonology).

Phonetics was substantially developed by the ancient Indians andby the medieval Arabs, but the modern tradition began in the six-

teenth century in England, and it was in nineteenth- and twentieth-

century Britain that such figures as Alexander Melville Bell, Henry

Sweet and Daniel Jones chiefly created modern phonetics, though

most of the instrumental techniques are far more recent. Phonetics

lies at the heart of dialectology.

See also: consonant; International Phonetic Alphabet; phonology; socio-linguistics; speech sound; transcription; vowel

Further reading: Ashby 1995; Ball and Rahilly 1999; Collins and Mees2003; Johnson 1997; Ladefoged 1993, 1971; Ladefoged and Maddieson1996; Laver 1994.

PHONOLOGY

The sound systems of languages, or the branch of linguistics which

studies these. Whereas phonetics is chiefly concerned with the phy-

sical nature of speech sounds, phonology deals with the ways in

which sounds behave in languages.

It took a long time for linguists to understand the difference

between phonetics and phonology, but, by the late nineteenth cen-

tury, the Polish linguists Miko�aj Kruszewski and Jan Baudouin deCourtenay had laid the foundations of phonology as a discipline.

The new phonological ideas spread out gradually cross Europe in

the early twentieth century; meanwhile, American linguists were

making similar progress somewhat independently.

The central concept in the new phonological approach was the

phoneme principle, an understanding of which permitted great

PHONOLOGY

216

advances in the analysis of the sound systems of languages. This

principle at last allowed linguists to understand the sounds of a

language as constituting an orderly system, instead of being a mere

collection of individual sounds; this insight was one of the earlysuccesses of the general approach to language study called

structuralism.

Important contributions to phonology were made by the Eur-

opean linguists of the Prague School in the 1930s and by the

American structuralists in the 1940s and 1950s. In the late 1950s,

phonology was transformed by the introduction of distinctive fea-

tures (phonological units smaller than phonemes); these features

were combined with certain ideas taken from Noam Chomsky’snew theory of transformational grammar to produce a dramatically

new approach named generative phonology, which focused on the

phonological process occurring in languages. Since the 1980s, gen-

erative phonology has broken up into a variety of more elaborate

and competing approaches, the majority of which are collectively

known as non-linear phonology; the two most prominent of these

are metrical phonology and autosegmental phonology, but several

others exist.

See also: minimal pair; neutralization; phoneme; phonetics; phonotactics

Further reading: Carr 1993; Collins and Mees 2003; Gussenhoven andJacobs 1989; Gussmann 2002; Hawkins 1984; Lass 1984; McCarthy2004, Shockey 2003, Sommerstein 1977.

PHONOTACTICS

The rules for combining phonemes into words in a language. Every

variety of every language possesses a larger or smaller set of pho-

nemes, and every legitimate word in that language must consist of a

permitted sequence of those phonemes. But the key word here is

‘permitted’: no language allows its phonemes to occur in just any

sequence at all. Instead, each language imposes strict limits on thesequences of phonemes which are allowed to occur in a word, and

those restrictions are its phonotactics.For example, English allows a word to begin with /b/ (as in bed),

with /r/ (as in red), with /l/ (as in led), with /n/ (as in net), with the

cluster /br/ (as in bread), and with the cluster /bl/ (as in bled). But it

does not permit a word to begin with the cluster /bn/: no such word

PHONOTACTICS

217

as *bned is even conceivable in English (the asterisk indicates this).

Moreover, if a word begins with /br/ or /bl/, then the next phoneme

must be a vowel: nothing like *blsed or *brved is possible either (but

such a combination may be possible in another language).Phonotactic constraints may differ widely from one language to

another, even when the sets of phonemes are somewhat similar. For

example, Hawaiian allows no consonant clusters at all, and every

syllable must end in a vowel, and so kanaka ‘man’ is a legal word,

but something like *kanak or *kanka is not. Japanese has a similar

rule, so when words are borrowed from English that would be ille-

gitimate phonotactically, they are nipponized: biru ‘beer’, sunugur-

asu ‘sunglasses’, gurufurendu ‘girlfriend’. The Caucasian languageGeorgian permits astounding consonant clusters, as in mts’vrtneli

‘trainer’ and vprtskvni ‘I’m peeling it’. The Canadian language Bella

Coola, unusually, permits words containing no vowels at all, like

�k’wtvw ‘make it big’!

See also: phoneme; phonology

Further reading: Hawkins 1984; Kreidler 1989.

PHRASE

A grammatical unit which is smaller than a clause. The term phraseis an ancient one, and it has long been used to denote a gramma-

tical unit which typically (though not invariably) consists of two or

more words, but which does not contain all of the things found in aclause. For example, a prepositional phrase consists of a preposition

with its object, as in under the bed, with her girlfriend and of the

wine.

Especially since the 1940s, linguists have recognized a much

larger variety of phrases than was formerly the case; among the

more important are the noun phrase (such as the little girl), the verb

phrase (such as was singing in the bath) and the adjective phrase

(such as pretty as a picture). Each of these types of phrase repre-sents a single phrasal category, and a phrasal category is one type of

syntactic category – that is, it represents one of the basic building

blocks used in constructing sentences. Generally speaking, at any

point in a sentence in which, say, a noun phrase can occur, any

noun phrase can be used, subject only to the (non-grammatical)

requirement of making sense.

PHRASE

218

The precise set of phrasal categories recognized varies some-

what according to the particular grammatical framework being

used; in particular, Government-and-Binding Theory recognizes a

number of phrasal categories not recognized by most other frame-works.

Observe that a phrase may consist of only a single word; for

example, the sentence Susie smiled consists of the noun phrase

Susie and the verb phrase smiled. These units are phrases because

they occupy the positions of phrases and behave like phrases;

strictly speaking, therefore, we have here a verb phrase smiled which

happens to consist only of the verb smiled. This double layer of

structure is routinely displayed in a tree diagram, and failure tomake it leads quickly to confusion.

Every phrase is built up from a head, an item (usually a single

word) which itself determines what kind of phrase the whole

thing is. A one-word phrase consists of a head with no other

material at all.

Occasionally, in modern linguistics, the term phrase is general-

ized to denote any syntactic unit in a sentence (any constituent), ofwhatever size or nature; this is the sense of the term in names likephrase-structure grammar. So, for example, in the sentence Susie

discovered that she was pregnant, the unit she was pregnant is a

complete clause, and hence it is not a phrase in the narrower sense

of the term, but it is a phrase in this extended sense.

See also: head; modifier


PHRASE-STRUCTURE GRAMMAR

A type of generative grammar which represents constituent structure

directly. We normally regard the structure of any sentence as an

instance of constituent structure, in which smaller syntactic units are

combined into larger units, which are then combined into stilllarger units, and so on. This kind of structure can be readily han-

dled by a phrase-structure grammar (PSG) (the full form of the

name is context-free phrase-structure grammar, or CF-PSG).

The idea of a PSG is simple. We first note what syntactic cate-

gories appear to exist in a given language, and what different inter-

nal structures each of these can have. Then, for each such structure,


219

we write a rule that displays that structure. So, for example, an

English sentence typically consists of a noun phrase followed by a

verb phrase (as in My sister bought a car), and we therefore write a

phrase-structure rule as follows: S ! NP VP. This says that a sen-tence may consist of a noun phrase followed by a verb phrase.

Further, an English NP may consist of a determiner (like the or my)

followed by an N-bar (like little girl or box of chocolates), and so we

write another rule: NP ! Det N0. We continue in this way until we

have a rule for every structure in the language.

Now the set of rules can be used to generate sentences. Starting

with S (for ‘sentence’), we apply some suitable rule to tell us what

units the sentence consists of, and then to each of those units weapply a further rule to tell us what units it consists of, and so on,

until we reach the level of individual words, at which point we

simply insert words belonging to the appropriate parts of speech.

The result is usually displayed graphically in a tree.

PSGs are the most appropriate type of grammar for teaching

elementary syntax, and moreover they are powerful enough to

describe successfully almost every construction occurring in any

language, though there exist one or two rare and unusual con-structions which they cannot handle.

PSGs were introduced by the American linguist Noam Chomsky

in the 1950s, but Chomsky had little interest in them, and they were

not really explored seriously until the British linguist Gerald

Gazdar began developing a sophisticated version around 1980; his

version was dubbed Generalized Phrase-Structure Grammar(GPSG). More recently, the Americans Carl Pollard and Ivan Sag

have constructed a very different-looking version called Head-Driven Phrase-Structure Grammar (HPSG), which is both linguisti-

cally interesting and convenient for computational purposes. HPSG

bases its formal rules on a richly organized and structured lexicon.

A lexical item (a word) is made of two features: its sound and its

syntactic-semantic information, which are in turn divided into fur-

ther sub-features. An example would be the verb ‘set in’ which

almost always has a semantics of negativity and adversity (bad

weather set in, financial problems set in) and cannot be used withhuman agents. To describe the grammar of the verb fully we need

to have a lexico-semantic information module built in to the

description. HPSG is an early example of the fact that grammar

and lexis are interdependent.

See also: constituent structure; generative grammar; tree


220

Further reading: Bennett 1995; Borsley 1996; Brown and Miller 1991;Lyons 1991; Pollard and Sag 1994.

PIDGIN

An auxiliary language created by people with no language in

common. Very many times in human history, people with no lan-

guage in common have found themselves thrown together and

obliged to deal with one another. Sometimes the language of just

one group will be learned by the others and used as a lingua franca,

but often something quite different happens: words from one ormore of the languages of the people involved will be taken and

stitched together into a kind of crude way of communicating. This

is a pidgin. A pidgin is nobody’s mother tongue, and it is not a real

language at all: it has no eleborate grammar, it is very limited in

what it can convey, and different people speak it differently. Still,

for simple purposes, it does work, and often everybody in the area

learns to handle it.

Pidgins can and do arise whenever the conditions are favourable,and very many have been created just in the past few hundred years.

Several were created along the east and west coasts of Africa, to

facilitate trading among Africans, Europeans and Arabs. Many

others were constructed in North America and the Caribbean,

particularly to enable African slaves to talk to one another and to

their European masters. And still others were constructed in the

Far East, mainly for trading purposes. The sugar plantations

of Hawaii attracted workers from a dozen countries in Asia andthe Pacific, and this led to the creation of yet another pidgin.

Many pidgins arise temporarily and disappear when the situation

around them changes: for example, during the USA–Korean War

(1950–3).

There are several possible fates for a pidgin. First, it may even-

tually drop out of use. This has happened to Hawaiian pidgin, now

almost entirely displaced by English, the prestige language of

Hawaii. Second, it can remain in use for generations, or even cen-turies, as has happened with some west African pidgins. Third, and

most dramatically, it can be turned into a mother tongue. This

happens when the children in a community have nothing but a

pidgin to use with other children, in which case the children take

the pidgin and turn it into a real language, by fixing and elaborat-

ing the grammar and greatly expanding the vocabulary. The result

PIDGIN

221

is a creole, and the children who create it are the first native

speakers of the creole.

See also: bioprogram hypothesis; creole; natural-language processing

Further reading: Chaudenson 2001; Holm 1988–9; Holmes 1992; Kou-wenberg and Singler 2005; Romaine 1994, 1988; Sebba 1997.

PLACE OF ARTICULATION

A label for the speech organs most directly involved in producing aconsonant. By definition, the production of a consonant involves a

constriction (narrowing or closure) somewhere in the vocal tract

between the glottis and the lips. We have a standard terminology

for labelling the particular parts of the vocal tract directly involved

in making that constriction, and each such label denotes a parti-

cular place of articulation.Toward the bottom end of the vocal tract, we can safely use

simple labels like glottal and pharyngeal. Farther up, we need inprinciple to identify both the lower articulator and the upper one

(in that order), and for this purpose we use a compound label

with the first identifier ending in -o and the second in -al (or -ar).

Examples:

� dorso-velar – back (dorsum) of tongue plus velum; e.g., [k]

� lamino-alveolar – blade (lamina) of tongue plus alveolar ridge;

e.g., English [n] for most speakers� apico-dental – tip (apex) of tongue plus upper teeth; e.g., French [t]

� sublamino-prepalatal – underside of tongue plus front of palate;

e.g., [”] in many Australian languages.

If the lower articulator is obvious or unimportant, we can omit it

from the label; hence we can say velar instead of dorso-velar, or

alveolar for any consonant involving the alveolar ridge.

A few traditional terms are unsystematic, such as retroflex forany consonant in which the tip of the tongue is curled up, palato-alveolar for a consonant involving a long constriction from the

alveolar ridge to the palate, and bilabial in place of the expected

labio-labial. Note also coronal for any consonant during which the

blade of the tongue is raised, whether or not the blade is involved in

the articulation.

PLACE OF ARTICULATION

222

For a consonant involving two simultaneous constrictions, we use

the -al ending twice, so that [w], for example, is labial-velar (thoughthe unsystematic labio-velar is also found).

See also: consonant; manner of articulation

Further reading: Ashby 1995; Ball and Rahilly 1999; Collins and Mees2003; Ladefoged 1971.

POLITENESS

The linguistic expression of courtesy and social position. While

politeness has non-linguistic aspects, we are here concerned only

with its linguistic expression. Except when we are deliberately

looking for a confrontation, we normally take care to ensure that

what we say (and what we don’t say) is chosen appropriately so as

to avoid embarrassing or offending anyone.

Sociolinguists often discuss politeness phenomena in terms of

face. Face is what you lose when you are embarrassed or humiliatedin public. We may distinguish your positive face (your need to

maintain and demonstrate your membership in a social group)

from your negative face (your need to be individual and indepen-

dent, to get what you want without offending anyone). A face-threatening act is any piece of behaviour which can easily make

another person lose face; a face-saving act is any piece of behaviour

which lessens or removes the threat of losing face.

The linguistic aspects of politeness have been much studied inrecent years, and a number of important variables have been iden-

tified: tone of voice, markers of status, terms of address, degrees of

certainty or confidence, discourse markers (like English please), the

choice between speaking and remaining silent, acceptability of

direct questions, and others. The rules of politeness vary con-

siderably from society to society, and it is very easy to give inad-

vertent offence when talking to speakers of another language.

For example, speakers of Malagasy (in Madagascar) consider itimpolite to give direct answers to questions or to make predictions

that might turn out to be wrong. Speakers of Navaho (in the USA)

consider it impolite to speak at all in the presence of a higher-

ranking person, or to provide their own names. Both Javanese and

Japanese have rich and complex systems for the overt linguistic

marking of status among speaker, listener and person talked about,

POLITENESS

223

including both different vocabulary and different grammatical

forms, and failing to mark status appropriately is a grave breach of

decorum. Note that it is inappropriate to say that X language is

more polite than Y language. Different languages simply encodetheir normal level of politeness with different features.

See also: communicative competence; ethnography of communication

Further reading: Bonvillain 1993; Brown and Levinson 1987; Eelen2001; Holmes 1992; Mills 2003; Watts et al. 2005.

POLYSYNTHESIS

The phenomenon in which a single word consists of more than one

morpheme. Languages which tend towards a 1:1 word to mor-

pheme ratio (such as Thai, Vietnamese and Bulgarian) are known

by contrast as analytic languages. The term polysynthetic is usually

reserved for those languages at the other extreme end of the scale,

which employ many morphemes (over 5:1) in word-formation.Native North American and Arctic languages are particularly well

known for their polysynthetic aspects. For example, the Siberian

Yupik language can express ‘He wants to get a big boat’ as angya-

rla-ng-yoocle-tuk (‘boat-big-get-wants-he’), where every morpheme

after angya ‘boat’ is a bound affix, and each of these bears no

relation to the full lexical word that could have been used in its

place. English has a few examples of long polysynthesis, mainly

technical terms from various fields and none in colloquial usage:diaminoethanetetraacetic acid, antidisestablishmentarianism, deoxy-

ribonucleic acid (DNA).

Polysynthesis is sometimes also used broadly to include examples

of incorporation, where compounding is made not from bound

morphemes but from full lexical items. Again, this too is common

in Native North American languages (which may account for the

confusion in usage of the term). English has an element of this:

breastfeed, dog-walk, self-harm, pub-crawl, overrun. Even morewidely, incorporation is sometimes used to include the process of

creating compound verbs.

See also: derivation; morpheme

Further reading: Baker 1996; van Valin and LaPolla 1997.

POLYSYNTHESIS

224

POSSIBLE WORLDS THEORY

Originally developed in the philosophy of language, this was a means

of determining the truth or falsity of propositions. A sentence couldbe judged either true, false or lacking a truth-value only when it

was considered to be uttered in relation to a set of states of affairs

that constituted the world surrounding the sentence. Our own rea-

lity is the actual possible world which could be considered alongside

other counterfactual possible worlds, in which sentences would have

different truth values. For example, This book is open about two-

thirds of the way through contains a proposition that is true if you

are reading the paper version of this book. However, if you arereading this as an e-book, then it is arguably either false or lacking

in truth-value.

Possible worlds theory, in this original sense, was a logical theory

and the worlds involved were simply reduced semantic sets. How-

ever, more recently, principles in possible worlds theory have been

developed in cognitive linguistics to be able to consider the sorts of

rich worlds that exist in mental representations. Work in schema

theory and text world theory, in particular, owes much to possibleworlds theory. These appraoches have all proven fruitful in the field

of cognitive poetics, a development of stylistics which uses cognitive

linguistics in the exploration of literary works.

See also: cognitive linguistics; schema; stylistics; text world theory

Further reading: Ronen 1994; Ryan 1991; Semino 1997; Stockwell2002b.

POWER

The relation between the type of language used by an individual

and that individual’s access to positions of power and influence. In

almost every society of any size or complexity, there are notable

differences in the type of language used by different sectors ofsociety, and, furthermore, associated differences in the access

enjoyed by the different groups to powerful and well-paid positions.

For example, in the English-speaking world, the contrast is pri-

marily between standard English and the various non-standard

forms of English, and it is obvious that speakers of standard Eng-

lish typically enjoy more power, more prestige and more money

POWER

225

than the others, who are far more likely to be confined to positions

of low prestige, with comparatively little money or influence and

often with little chance for advancement. What, if anything, should

we do about this state of affairs?Broadly speaking, there have been two answers proposed to this

question. Some people emphasize the importance of teaching pres-

tige forms of language to speakers of non-prestige forms in order to

empower them, while others object to the imposition of prestige

forms on the ground that it effectively denies power to speakers of

other varieties. In the English case, the first group are essentially

arguing that it is the very command of standard English which

itself confers access to power and prestige and, hence, that it is oursocial duty to extend this command to as many people as possible,

since otherwise non-standard speakers will remain marginalized

and unable to play a full part in society. The second group, in great

contrast, argue that standard English is nothing more than the

identifying badge of a particular and maximally powerful group, a

mere sign of class membership and, hence, that stressing the impo-

sition of standard English denies the value of non-standard forms,

so that non-standard speakers are implicitly dismissed as inade-quate and unworthy of power. Many members of the second group

go so far as to demand that non-standard varieties of English

should be formally recognized as the equal of standard English,

and that non-standard English should become the vehicle of edu-

cation for its speakers, a position which appals the members of the

first group.

Much commentary on language by journalists, politicians and

other non-linguist observers is primarily concerned with power andauthority in language, even where this is not explicit.


Further reading: Crowley 1989; Fairclough 2001; Honey 1997.

PRAGMATICS

The branch of linguistics which studies how utterances commu-

nicate meaning in context. The study of meaning, commonly known

as semantics, has long been one of the most daunting and difficult

areas of language study. In the 1950s and 1960s, however, linguists

and philosophers slowly began to realize that part of the difficulty

PRAGMATICS

226

lay in their failure to distinguish two quite different aspects of

meaning.

The first type of meaning is intrinsic to a linguistic expression

containing it, and it cannot be separated from that expression. Thestudy of this kind of meaning is the domain of semantics, as we

now understand the term. But there is a second kind of meaning,

one which is not intrinsic to the linguistic expression carrying it,

but which rather results from the interaction of the linguistic

expression with the context in which is it used. And to the study of

this kind of meaning we give a new name: pragmatics.Consider the sentence Susie is a heavy smoker. In all circum-

stances, this sentence carries with it its intrinsic meaning: Susiesmokes a large quantity of tobacco every day. This meaning is

intrinsic and inseparable. But now consider what happens when this

sentence is uttered as a response to three different utterances pro-

duced by Jessica in three different contexts.

1 [Jessica is trying to have smoking banned in offices]: Can you ask

Susie to sign this petition?

2 [Jessica is trying to arrange a blind date for Dave, a non-smokerwho hates cigarette smoke]: Would Susie like to go out with

Dave?

3 [Jessica, a medical researcher, is looking for smokers to take part

in some medical tests]: Do you know of anybody I could ask?

In each case, you will agree, something very different is being

communicated. In the first case: Susie is unlikely to sign the peti-

tion, so there’s no point in asking her. In the second: Dave andSusie won’t get on, so there’s no point in fixing them up. Third:

Susie will be a suitable person for your study.

Now, it is not possible to maintain that this single unvarying

sentence actually means all of these different things. Rather, these three

meanings have been communicated as a consequence of the inter-

action between what was said and the context in which it was said.

Every time the context changes, what is communicated changes as well.

And it is this variable, context-bound relation between what is saidand what is communicated that is the subject-matter of pragmatics.

It should be noted that, outside the USA, the term pragmatics is

often used in a much broader sense, so as to include a great number

of phenomena that American linguists would regard as belonging

strictly to sociolinguistics: such as politeness, narrativity, and the

signalling of power relations.

PRAGMATICS

227

See also: conversational implicature; cooperative principle; meaning;relevance; semantics; speech act

Further reading: Blutner and Zeevat 2003; Grundy 1995; Schiffrin 1994;Thomas 1995; Verschueren 1999; Yule 1996.

PREDICATE

That part of a sentence other than its subject. It was the Greek

philosopher Aristotle who first divided sentences into subjects and

predicates. Given a fixed subject Susie, we can construct sentencesby adding to this subject any number of different predicates:

smokes, is clever, has been promoted, wants to buy a new car, believes

that astrology is garbage. In each case (in English), the predicate

role is filled by a verb phrase; this is typically so in most languages,

though some languages permit predicates belonging to other syn-

tactic categories.

It is important to realize that logicians use the term predicate in a

very different way, one which is now prominent in linguistics, espe-cially in semantics. In the system of formal logic called predicate logic,the sentence Bruce is bald would typically be represented as Bald

(Bruce). Here Bald is a (logical) predicate corresponding to English is

bald, and it is a one-place predicate requiring only one argument

(noun phrase) in order to be satisfied. The sentence Bruce loves Kathy

might be similarly represented with a one-place predicate as Love-

Kathy (Bruce), but more commonly we would represent it as with a

two-place predicate, as follows: Love (Bruce, Kathy). This time thepredicate Love requires two arguments in order to be satisfied.

Since both of these conflicting usages are widespread in linguistics,

you must take care that you understand which sense is intended

when you encounter the term.

See also: verb phrase


PREPOSITION

The part of speech which includes words like to, with and of. Like

many other languages, English possesses a smallish class of words

PREDICATE

228

called prepositions. A preposition has only one major property: it

combines with a following noun phrase – its object – to form a

larger syntactic unit – a prepositional phrase. Typical prepositionsinclude of, to, in, for, with, under, about, inside, after, in front of andin spite of. (Note that some of these can also belong to other parts

of speech.) Here are some typical prepositional phrase: to the car,

for a while, with Susie, after the war, in spite of the weather. Some

prepositions, like under and after, express identifiable meanings.

Others, such as of, have a purely grammatical function: in the noun

phrase the end of the year, the preposition of serves merely to con-

nect the smaller noun phrase the year to the rest of the bigger one,

but of the year is still a prepositional phrase.Prepositions are analysed in cognitive linguistics as the realization

of image schemas: basic representations of the physical and spatial

movement involved in through, over, under, into, from, beside and so on.

In some languages, such as Japanese and Basque, the items that

do the job of prepositions follow their object instead of preceding

it, and we therefore call them postpositions. Japanese examples:

Tokyo ni ‘to Tokyo’; Tokyo de ‘in Tokyo’. Prepositions and post-

positions together are collectively called adpositions.Some languages, such as the Australian languages, lack adposi-

tions altogether, and use entirely different grammatical devices to

do the same job.

See also: landmark; syntax; trajector

Further reading: Collins Cobuild 1990; Hurford 1994.

PRESCRIPTIVISM

The imposition of arbitrary norms upon a language, often in defi-

ance of normal usage. Every language exhibits a good deal of

regional and social variation. If very many people want to use a

language for a number of different purposes, then it is convenient

and even necessary to have a single agreed form of the language – astandard language – known and used by everybody, or at least by all

educated speakers. Otherwise, if people insist on using their own

particular varieties, the result will be confusion and mis-

understanding. But, since languages are always changing, there will

always be doubts and disagreements over which forms and usages

should be recognized as part of the standard language.

PRESCRIPTIVISM

229

Prescriptivism consists of the attempts, largely by teachers and

writers, to settle these disagreements by insisting upon the use of

those particular forms and usages which they personally prefer and

by condemning those others which they personally dislike. Ofcourse, some degree of prescriptivism is necessary, particularly in

education: people who naturally use forms which are blatantly not

accepted as standard by the community as a whole must learn to

use the standard forms, at least in those circumstances which

call for the standard language, or else they will be severely dis-

advantaged.

But the problem is that many prescriptivists go much further

than this convenient codification, and try to condemn usages whichare in fact perfectly normal for even educated speakers, and to

insist instead upon usages which were current generations or cen-

turies ago but which are now effectively dead, or even upon usages

which have never been normal for anybody.

A famous example concerns the so-called split infinitive. For

generations, virtually all English-speakers have spontaneously said

things like She decided to gradually get rid of the teddy-bears she had

spent twenty years collecting. Here the sequence to gradually get rid

of is the ‘split infinitive’. Many prescriptivists have condemned this

usage, on the supposed ground that to get is a single verb-form, the

‘infinitive’, and therefore ‘logically’ cannot be split up. Such people

typically insist instead on something like She decided gradually to

get rid of . . .. But this is all wrong.First, the proposed ‘correction’ is badly misleading: it suggests

that it is the decision which is gradual, rather than the disposal.

Second, the sequence to get is not an infinitive, nor is it a verb-form, nor is it even a grammatical unit at all. The true infinitive

here is get, while to is nothing but a linking particle. The adverb

gradually logically belongs next to get rid of, and that’s where

speakers normally put it. That to get is not a grammatical unit can

be shown in a number of ways, not least of which is the observation

that speakers regularly break it up. (Another test is the construction

illustrated by She has asked me to change my hairstyle, but I don’t

want to, in which the understood change is deleted while to isobliged to remain – hardly possible if to change were really a unit.)

Hence the prescriptivists’ position is ignorant and wrong-headed: it

represents an attempt to replace normal and elegant usage by

something which is silly, unnatural and hard to understand, and

which is used by nobody except some prescriptivists and those few

who take them seriously.

PRESCRIPTIVISM

230

Many prescriptivists also object to the familiar English practice

of ending a sentence with a preposition, apparently on the bizarre

ground that this construction is not possible in Latin. They take

exception to ordinary English utterances like Who were you talking

to?, What’s this gadget for? and That’s something I just can’t put up

with, demanding instead unnatural things like To whom were you

talking?, For what is this gadget?, and I have no idea what they

would do about the last one.

Prescriptivists also reject such ordinary utterances as Who do you

trust?, demanding instead Whom do you trust?, a form which was

current hundreds of years ago but is now dead, except in frostily

formal styles of speech and writing. Other examples of feeble-mindedness include the aversion to multiple negation (double

negatives for emphasis); prejudice against ain’t; insisting on fewer

for count-nouns instead of less (Five items or less at the super-

market check-out); distaste for the impersonal you for one (one

doesn’t do such things); and hang-ups over spellings and apos-

trophes when there is no communicative problem at all. It is deeply

unfortunate that so many commentators have seen fit to lose touch

with reality and to pursue their own absurd little bugbears at suchlength and with such passion.

See also: descriptivism; purism

Further reading: Crystal 1997; Edwards 1994; Milroy and Milroy 1997;Pinker 1994.

PRESTIGE

A feature identified in sociolinguistics as the high esteem and social

value which speakers attach to certain language varieties. For

example, in Britain the received pronunciation (RP) accent is

regarded as prestigious even though it is spoken by less than 5 per

cent of the UK population: it is seen as an indicator of education,

wealth, trustworthiness, attractive looks and honesty! The oppositeof prestige is stigma, which attaches to other varieties depending on

the viewpoint of the stigmatizer. Some varieties display covert pres-tige, where speakers will share a sense of value for the standard

language, but will unconsciously accentuate their own stigmatized

variety in certain contexts. For example, middle-class men in var-

ious parts of England have been observed switching away from RP

PRESTIGE

231

and towards the local working-class accent when the setting is per-

ceived as being casual, colloquial and male-oriented. They will

often hypercorrect (i.e., overdo) the working-class accent in this

situation. The effects of prestige, stigmatization and hypercorrectionare important factors in language change.

See also: accent; sociolinguistics

Further reading: Bex and Watts 1999; Llamas et al. 2007; Trudgill 1995.

PRESUPPOSITION

A particular sort of inference. Consider the sentence John’s wife

runs a boutique. Now, when we hear this, we are immediately enti-

tled to draw the following inference: John is married. This is an

example of a presupposition: we say that the first sentence pre-supposes the second.

A presupposition differs from other types of inference, such as

an entailment or a conversational implicature, in several ways. Mostobviously, a presupposition survives negation. If I negate the ori-

ginal sentence, the result is John’s wife doesn’t run a boutique – and

this still presupposes that John is married. Both the original sen-

tence and its negation have the interesting property that they are

bizarre and uncooperative things to say if John is in fact a bache-

lor. Very informally, then, a presupposition of statement P is

something which has to be true before P can possibly be a plausible

thing to say.In a widespread view, introduced by the British philosopher Peter

Strawson around 1950, we can say that P presupposes Q if and only

if Q has to be true before P can be either true or false. That is, if Q

is false, then P can be neither true nor false: it simply has no truth

value at all. So, if John is a bachelor, then John’s wife runs a bou-

tique is not false, nor is it true; it is simply devoid of truth value.

But not everyone agrees with this interpretation, and some linguists,

following the British philosopher Bertrand Russell, would prefer toconclude that the statement really is false. This disagreement is

bound up in a larger controversy over whether presuppositions

properly belong to the domain of semantics or to that of

pragmatics.

Note that questions and commands can have presuppositions.

The question Do you still drink a bottle of vodka every night? is

PRESUPPOSITION

232

bizarre and unanswerable if you have never drunk vodka on that

scale (this is what is informally known as a ‘loaded question’), and

the command Take your hat off! is bizarre and impossible to obey if

you aren’t wearing a hat.

See also: conversational implicature; entailment

Further reading: Grundy 1995; Levinson 1983; Saeed 1997; Verschueren1999.

PRODUCTIVITY

The degree of freedom with which a particular grammatical pattern

can be extended to new cases. We most often speak of productivity

in connection with patterns of word-formation. The noun-forming

suffix -ness is highly productive: happiness, preparedness, salacious-

ness, user-friendliness. The same is true of the verbal prefix re-:

rewrite, reconsider, reappoint, renegotiate, reboot (a computer).

But the noun-forming suffix -th is totally unproductive: we haveexisting cases like warmth and depth, but we cannot form any new

ones: *happyth, *bigth, *sexyth (the asterisk indicates forms that are

unacceptable). The noun-forming suffix -dom is weakly productive:

to established cases like kingdom and martyrdom we occasionally

add new ones like gangsterdom, tigerdom and stardom, but we can’t

do this freely: *policedom, *universitydom, *childdom.

The adverb-forming suffix -wise was formerly unproductive and

confined to a few cases like clockwise and otherwise, but today wefreely coin new formations like healthwise, moneywise, clotheswise

and fitnesswise. The noun-prefix mini- didn’t even exist before 1960,

but today it is prodigiously productive: miniskirt, minicomputer,

mini-microphone, minibus, even mini-war.

See also: word-formation

Further reading: Bauer 1988; Katamba 1994; Matthews 1991.

PRONOUN

The part of speech which includes words like she, them and some-

thing. Pronouns have been recognized as a distinct part of speech

PRONOUN

233

since ancient times. Essentially, a pronoun is a single word (or

rarely a longer form), with little or no meaning of its own, which

functions as a complete noun phrase.

Pronouns are classified into several types. A personal pronounpoints at some participant in a speech situation: I, you, we, she,

they. A demonstrative pronoun points in space or time, like the this

in This is a good book. An interrogative pronoun asks a question,

like who in Who’s there? An indefinite pronoun, such as somebody or

anything, fills a slot in a sentence without providing much specific

meaning, as in Do you need anything? A relative pronoun introduces

a relative clause, like the who in The students who streaked the gra-

duation ceremony are in trouble. Finally, a reflexive pronoun likeherself and a reciprocal pronoun like each other refer to other noun

phrases in the sentence in specific ways, as in She cursed herself and

They are seeing a lot of each other.

As a rule, a pronoun cannot take a modifier, but there are a few

exceptions: little me, poor you, something interesting.

See also: anaphor; noun phrase

Further reading: Collins Cobuild 1990; Hurford 1994; Jackson 2002;Jeffries 2006.

PROSODY

Variations in pitch, loudness, rhythm and tempo (rate of speaking)

during speech. The term prosody is an ancient one, but it was ori-

ginally applied only to the analysis of verse structure. Linguists in thetwentieth century have taken over the term and applied it specifically

to the several variations in the behaviour of the voice cited above.

These prosodic features may or may not be expressly linguistic in

nature. Tone languages and languages with pitch or stress accents

make specific linguistic use of pitch and/or loudness, but utterances

in all languages are characterized by noticeable variations in all

four of these features. Particularly important is the use of pitch

variations in intonation, which is universal in languages.Prosodic phenomena are notoriously difficult to study, but some

considerable progress has been made in examining at least some of

them. Nevertheless, elementary textbooks of phonetics and of pho-

nology only rarely discuss prosodic features in any detail.

It should be noted that the terms prosody and prosodic also have

some rather special uses. In the theory of phonology called prosodic

PROSODY

234

phonology, the term prosody is given a highly distinctive technical

sense which is peculiar to that theory. In some contemporary the-

ories of phonology and of morphology, the adjective prosodic is

likewise given some rather distinctive senses for labelling thingswhich are important in those theories. Finally, note that in some

quarters prosody is used merely as a synonym for suprasegmental, a

usage which is not recommended.

See also: intonation; stress; suprasegmental; tone language

Further reading: Ball and Rahilly 1999; Laver 1994; Shockey 2003.

PROTO-LANGUAGE

The hypothetical ancestor of a language family. When we find some

languages which are clearly connected in a genetic relationship, and

which therefore form a language family, it follows by definition that

all are descended from a single ancestral language; that is, that they

all started off long ago as no more than regional dialects of thatancestral language. Given enough data, historical linguists can

apply comparative reconstruction to obtain substantial information

about what that ancestral language was like, even though in most

cases it was never itself recorded. The reconstruction which they

obtain in this way is a reasonably accurate picture of the ancestral

language, which is called the proto-language for the whole family.

A proto-language is named by prefixing Proto- to the name of

the family. For example, the ancestor of the Germanic languages(English, German, Swedish and others) is Proto-Germanic; the

ancestor of the Romance languages (Spanish, French, Italian and

others) is Proto-Romance; and the ancestor of the vast Indo-

European family is Proto-Indo-European.

See also: Indo-European; language family; reconstruction

Further reading: Hock 1986; Trask 1996.

PROTO-LANGUAGE HYPOTHESIS

The hypothesis that human language arose rather abruptly from a

vastly simpler precursor. The proto-language hypothesis was developed

PROTO-LANGUAGE HYPOTHESIS

235

in the 1980s by the British-born American linguist Derek Bick-

erton. A keen student of pidgins and creoles, Bickerton was

impressed by the rapidity with which a pidgin (which is not a full

language) can be converted into a creole (which is) when the con-ditions are right. He therefore developed the argument that some of

our remote ancestors must have had only proto-language, a crude

and limited system somewhat resembling a pidgin, and that this

must have developed, probably very suddenly, into true language

when certain critical conditions were in place in the brain.

Bickerton argues that proto-language can still be observed today

in certain circumstances: in pidgins, in the speech of individuals

suffering from certain kinds of disability, in the speech of veryyoung children, and in the severely limited linguistic accomplish-

ments of laboratory animals instructed in something resembling a

human language. The hypothesis is deeply controversial.

See also: animal communication; origin and evolution of language

Further reading: Bickerton 1981, 1990, 1996.

PROTOTYPE

An important feature of the way that semantic knowledge is held in

memory, and therefore an important concept in semantics, psycho-

linguistics and cognitive linguistics. Human categories have been

found to display prototype effects, so, for example, most people

have a sense of the ‘best example’ of a bird: sparrow, pigeon, crow,perhaps, rather than emu, ostrich, penguin or dromas ardeola (crab

plover). Similar ranges of examples can be applied to categories

such as furniture, fruit, linguists and in fact every conceivable cate-

gory, including provisional categories such as things to do on a

Friday night, what to wear to a family party, and so on. Categor-

ization is thus not an ‘either-or’ matter but a matter of radiality,with central examples of a thing and peripheral or poor examples

of a thing. Thus a potato is not simply a ‘non-fruit’; it is just a verybad example of a fruit. You can test this by noting that it is a better

example of a fruit than a television set.

Note that prototypical organization is culture specific. The crab

plover might be a very prominent example of a bird if you live around

the Indian Ocean. Prototypes are also malleable: a screwdriver can be

recategorized as a good example of an envelope-opener if it is all

PROTOTYPE

236

that you have to hand. Prototypicality seems to be the cognitive

mechanism underlying the notion of basic level concepts which tend

to be the most neutrally lexicalized (bird, dog, computer, rather than

starling, terrier or IBM).

See also: cognitive linguistics; semantics

Further reading: Lakoff 1987; Taylor 2003; Ungerer and Schmid 1996.

PSYCHOLINGUISTICS

The study of the connections between language and mind. Psycho-

linguistics began to emerge as a distinct discipline in the 1950s. To

some extent, its emergence was promoted by the insistence at the

time of the linguist Noam Chomsky that linguistics should be

regarded as a part of cognitive psychology, but there were other

factors as well, notably the growing interest in the question of lan-

guage acquisition by children.

There is no doubt that the study of acquisition has so far beenthe most prominent and successful area of psycholinguistics. But a

number of other topics have also been explored, with varying

degrees of success. Many of these are aspects of language proces-

sing, the steps involved in producing and comprehending speech.

Others include the links between language use and memory, the

linguistic examination of reading, and more recently possible links

with perception and cognition.

We now possess a great deal of data in most of these areas, butprogress in developing theoretical interpretations has been slow.

The enthusiastic early attempts at understanding mental processing

of language in terms of transformational grammar proved a failure,

and contemporary theorizing tends to be less ambitious: grand

schemes are out, and psycholinguists now content themselves with

trying to provide accounts of specific aspects of language behaviour.

Psycholinguists would also like to link their findings to those of

neurolinguistics, the study of language and brain, but this has notproved at all easy. Some psycholinguists are also contributing to the

development of cognitive linguistics and of cognitive science generally.

See also: language acquisition; language processing; neurolinguistics

Further reading: Aitchison 1989, 1994; Field 2003; Garman 1990;Garnham 1985; Steinberg 1993.

PSYCHOLINGUISTICS

237

PUNCTUATION

A conventional system of marks representing information about the

structure of a written text. The earliest writing systems used nopunctuation, and often they didn’t even separate words. The Greeks

introduced the earliest known punctuation during the classical

period, but it was very different from ours. In those days, and for

long after, silent reading was unknown, and the reader of a text

spoke the words out loud, even when reading alone.

Greek orators who were preparing speeches for delivery found it

helpful to add marks to their texts to remind them where to pause

briefly, where to make a dramatic pause, where to raise the voice,and so on. And these marks, added purely for rhetorical purposes,

were the first punctuation.

The Romans and the medieval Carolingians gradually elaborated

the set of punctuation marks, and these marks began to be used to

indicate structural aspects of texts, rather than breathing places.

With the eventual development of silent reading, this structural

function entirely displaced the old oratorical one, and today our

punctuation systems express mainly structural information.All other widely used writing systems have also developed sys-

tems of punctuation, though it is notable that a standard punctua-

tion system is generally the very last aspect of a writing system to

become established. Even in Europe, languages like English, French

and German, all written in the roman alphabet, all use slightly dif-

ferent punctuation and, moreover, American punctuation differs in

a few respects from British. Punctuation can also change: modern

English punctuation is very different from that used in the eight-eenth century.

See also: writing system; orthography

Further reading: Foley 1993; Sampson 1985.

PURISM

The belief that words (and other linguistic features) of foreign

origin are a kind of contamination sullying the purity of a lan-

guage. Almost all languages are in contact with other languages,

and all languages borrow words (and sometimes other features)

from those neighbouring languages. Sometimes speakers of the

PUNCTUATION

238

receiving language take exception to the presence in their language

of these ‘foreign’ elements, which they see as ‘impurities’, and agi-

tate to have them removed and replaced by native elements. This

attitude is called purism, and it is widespread.For example, the French authorities, dismayed by the hundreds

of English words pouring into French every year, have made stren-

uous efforts to eradicate some of these English words from their

language and to replace them with novel French creations, and

their recommendations have the force of law in certain domains.

Thus, computer has been officially replaced by ordinateur, software

by logiciel, light pen by crayon optique, floppy disc by disquette, pie

chart by camembert, videoclip by bande promo, bookmobile first bybibliotheque circulante but now by bibliobus, and bulldozer by bou-

ledozeur (!) (this last was a failure and has been abandoned). The

English word debugging was at first rendered as debogage, but more

recently the amusing deverminage has been coined. Often such

coinages fail under pressure from popular usage.

Purism has almost never been a force among speakers of English,

but speakers of French, German, Icelandic, Turkish and Basque

(just to name a few) have at times engaged in large-scale purges offoreign elements, with varying degrees of success.

See also: prescriptivism

Further reading: Milroy and Milroy 1997.

QUALITATIVE APPROACH

A trend towards the description and explanation of language use

within naturally occurring social and cultural settings. The various

strands of what we now call the qualitative approach have been

visible within sociology since the 1920s, but only since the 1980s has

this approach become prominent in the study of language.

This label is applied to a particular methodology, or set of

methodologies, particularly within sociolinguistics and conversationanalysis, in which authenticity is regarded as crucial, and in which

authenticity is achieved through the detailed description of the

social settings and social practices which yield language data. A

qualitative approach typically focuses on the study of small num-

bers of speakers or texts, since an abundance of data and statistical

studies are seen as less important than revealing the social meanings

QUALITATIVE APPROACH

239

which speakers and writers attach to their linguistic activities.

Consequently, this approach contrasts rather vividly with the

quantitative approach. (Moreover, the depth of detailed data col-

lected in qualitative research renders it almost impossible to dealwith more than a few informants.)

Qualitative research employs observation, textual analysis, inter-

views, and the recording and transcribing of speech. While these

may also be used in quantitative work, in qualitative research they

are given a different status. For example, in quantitative investiga-

tions, observation is limited to the initial exploratory stages, essen-

tially to the stages at which the topic of research is being identified,

since observation alone is here considered to be too unreliable toyield usable data. In qualitative research, however, the observation

of participants is considered fundamental in revealing the social

and political factors which underpin language use. Thus, disciplines

which regularly employ qualitative methodology, such as ethno-

graphy, social anthropology and cognitive anthropology, often rely

solely upon observation and other ‘open-ended’ research procedures.

See also: conversation analysis; ethnography of communication; quantitativeapproach

Further reading: Johnstone 2000; Milroy 1987b; Milroy and Gordon2003; Silverman 1993; Wengraf 2001.

QUANTITATIVE APPROACH

A statistical approach to the study of variation in language. Varia-

tion in the way people use their language is very prominent, but for

generations linguists could see no way of making any sense of it:

instead, variation was typically seen as an irrelevant nuisance and

ignored. In the 1960s, however, the American sociolinguist William

Labov pioneered a wholly new approach to language study, one

which made variation itself a prime object of examination. The key

to this approach was the introduction of statistics: Labov collectedstatistical data on the frequency of competing forms used by dif-

ferent speakers and then looked for correlations with non-linguistic

factors. This is the quantitative approach, and it has proved out-

standingly successful.

What Labov and his successors found is this. If we merely

observe a speaker, or a group of speakers, all we can notice is that

QUANTITATIVE APPROACH

240

speakers sometimes use this form and sometimes that one, in a

seemingly haphazard manner. However, if we tabulate the frequency

of each competing form, we often find first that one individual dif-

fers notably from another and second that the frequency of a par-ticular form correlates strongly with some non-linguistic variable.

Most often, the non-linguistic variables are obvious ones like sex,

age, social class and degree of formality, but sometimes we find

more unexpected correlations.

What these studies have shown is that variation is not haphazard

at all. Instead, variation is highly structured: statistically speaking,

each individual behaves in a highly consistent way, with some indi-

viduals behaving differently from others, depending on some of thevariables just mentioned, and, moreover, the behaviour of each

individual changes in a predictable way when the context of speak-

ing becomes more or less formal.

The quantitative approach has revolutionized the study of lan-

guage by demonstrating that linguistic behaviour is even more

highly structured than we had previously suspected; it has con-

tributed enormously to the study of language change, and it has

provided a resolution of the Saussurean paradox.

See also: language change; social stratification of language; sociolinguistics;variation

Further reading: Hudson 1996; Labov 1972; Milroy 1992.

RAISING

Any of various phenomena in which a linguistic element appears in

a higher clause than is semantically appropriate. Consider the sen-

tence It appears that Susie is falling asleep. Here Susie is, both logi-

cally and grammatically, the subject of the verb phrase is falling

asleep within the lower (subordinate) clause. But the same infor-

mation can be expressed differently: Susie appears to be falling

asleep. This time Susie, still the logical subject of be falling asleep,appears as the grammatical subject of appears. In this case, we say

that Susie has been raised out of the lower clause into the higher

one, and, since the item undergoing raising is the subject of the

lower clause, we call this phenomenon subject-raising, or sometimes

more specifically subject-to-subject raising, since the raised element

ends up being the subject of the higher clause.

RAISING

241

This is not the only kind of raising. Consider the example I

believe that she is happy. Here she is again the subject of the lower

clause. But we can express this as I believe her to be happy, in which

her is apparently now the object of the verb believe in the higherclause, as shown by its form. This is subject-to-object raising. (Anote: a minority of linguists would prefer to analyse this last

example as having the structure I believe [her to be happy], with her

still being the subject of to be happy; proponents of this unusual

analysis sometimes refer to the bracketed sequence as a smallclause.)

Now consider the example. It is difficult to please Susie, in which

Susie is the object of the verb please. This can equally be expressedas Susie is difficult to please, in which Susie has been raised to

become the subject of is difficult. This last construction was there-

fore formerly called object-to-subject raising, though today it is

more usually known by the quaint name tough-movement, since it ismost frequent with verb phrases expressing degrees of difficulty.

Finally, consider the example It seems that I never meet the right

women. This is more idiomatically expressed as I never seem to meet

the right women, in which the negative item never has been raisedout of the lower clause, where it logically belongs, into the higher

clause. This is negative raising, and it is illustrated also by cases like

I can’t seem to find my keys, equivalent to It seems that I can’t find

my keys.

Raising phenomena are complex and intricate, and they have

received a good deal of attention from syntacticians in recent years.

See also: control, movement

Further reading: Borsley 1991.

RANK SCALE

In systemic linguistics but also commonly in language description,

language is divided into several levels from the largest units downto the smallest, roughly as follows:

discourse

text

sentence

clause

RANK SCALE

242

phrase

lexeme

morpheme

phoneme

Each level corresponds roughly with a sub-discipline within lin-

guistics: discourse analysis, text linguistics, syntax and semantics,

lexicology, morphology, and phonology. In Systemic Linguistics, the

rank scale is hierarchical, with each level above encompassing the

levels below. So, for example, syntactic embeddedness can be

described in terms of elements from one level being rank-shifted to

another level: The man who wrote this book also wrote that book

contains a clause that is rank-shifted into the noun-phrase.

See also: Systemic Linguistics

Further reading: Halliday 2004; Thompson 2004.

RECONSTRUCTION

Working out features of dead and unrecorded languages, or of

unrecorded earlier stages of single languages. In the vast majority of

cases, a given language can be clearly seen to share a common

ancestry with at least some other languages – that is, the languages

in question all started out long ago as no more than dialects of a

single ancestral language, which itself is almost never recorded. In

such cases, practitioners of historical linguistics are naturally inter-ested in trying to work out as much as they can about the nature of

that unrecorded ancestral language.

Moreover, even a single language is only ever recorded since

some particular point in time, often a rather recent one, and his-

torical linguists are also interested in obtaining information about

the prehistory of that language.

Since the end of the eighteenth century, historical linguists have

been painstakingly constructing effective and reliable methods ofperforming these tasks, and we now know a great deal about how

to go about them. The business of obtaining information about

unrecorded languages is reconstruction, and there are two main

types.

In comparative reconstruction, a linguist compares several lan-

guages which are known to be related; each of these languages

RECONSTRUCTION

243

normally preserves some features of their common ancestor, but

has lost others. Since different languages preserve different ances-

tral features, it is often possible to determine in some detail what

the ancestral language was like, and hence what changes haveoccurred in each of the daughter languages. Key to this is the

identification of cognates: words, phrases or grammatical patterns

in two languages that have a common root.

In internal reconstruction, the linguist works with only a single

language and tries to determine what an earlier unrecorded stage of

that language was like, and hence what changes occurred to pro-

duce the earliest recorded form.

Both types of reconstruction are most commonly performedupon recorded languages, but, once we have managed to recon-

struct some unrecorded languages in reasonable detail, it is per-

fectly possible to apply the same methods to these and to

reconstruct even further back in time. Naturally, though, a recon-

struction based upon reconstructions is generally less secure than

one based upon attested languages, and in practice there are limits

on how far back we can go.

See also: comparative reconstruction; historical linguistics; Indo-European;internal reconstruction; language change

Further reading: Fox 1995; Hock 1986; Hock and Joseph 1996; Trask1996.

RECURSION

The occurrence in a sentence of a syntactic category containing

within it a smaller instance of the same category. Recursion is per-

vasive in the grammars of the languages of the world, and its pre-

sence is the chief reason we are able to produce a limitless variety of

sentences of unbounded length just by combining the same few

building blocks.

For example, a noun phrase (NP) in English may contain within ita prepositional phrase (PP), and a prepositional phrase always con-

tains a noun phrase. Hence we can build up an NP containing a PP

containing an NP containing a PP . . . and so on, as far as we like.

In the sentence I’ve bought a book about the history of the debate

between defenders of different theories of education, everything after

bought is a single large NP beginning with a book; this NP contains

RECURSION

244

a single large PP beginning with about, which in turn contains

another NP beginning with the history, which contains a smaller PP

beginning with of, and so on, until the final NP education contains

no further PP. Here both the NPs and the PPs illustrate recursion.Similarly, the well-known verse containing the line This is the dog

that chased the cat that killed the rat that ate the malt that lay in the

house that Jack built illustrates the recursion of relative clauses,

with each that introducing a new relative clause embedded within

the one already begun, so that everything after dog is one huge

relative clause containing a series of smaller relative clauses, each

one embedded within another.

See also: constituent structure; phrase; syntax

Further reading: Brown and Miller 1991; Keyser and Postal 1976.

REFERENCE

The relation between a linguistic expression and something which itpicks out in the real or conceptual world. A linguistic expression which

refers to (points at) something in the non-linguistic world is a referringexpression, and the most familar such expressions are noun phrases.

A simple noun like dog does not itself refer to anything: it merely

denotes the entire class of dogs. But, when it is embedded in a sui-

table noun phrase, the whole phrase may succeed in referring to

something specific in the world: this dog, the little dog with the

floppy ears, that dog that Aunt Sophie used to have, and so on. Andthe relation between such an expression and the thing it picks out is

one of reference.Proper names may also refer and, indeed, they hardly do any-

thing else: Abraham Lincoln, the Golden Gate Bridge, Spain – these

really have no intrinsic meaning at all, and they merely point to

particular entities.

Reference is entirely in the mind of the speaker, and the same

linguistic expression may be a referring expression sometimes butnot at other times. If I say Janet wants to marry a Norwegian, and I

have in mind a particular Norwegian who Janet wants to marry –

say, Olaf Thorqvist the Olympic skier – then a Norwegian is clearly

a referring expression (here a specific indefinite). But, if I say the

same thing, meaning only that Janet has a curious desire to acquire

a Norwegian husband, any Norwegian husband, then a Norwegian

REFERENCE

245

is not a referring expression in this strict sense (it is a non-specificindefinite), since I have nobody in mind for the linguistic expression

to point at.

There is a further complication with expressions like the fastest

runner in Brazil. I can use this quite happily without having the

faintest idea who that fastest runner in Brazil might be. Here I am

merely assuming that the expression must, in principle, pick out

someone or other, even though I don’t know who. Such an expres-

sion is an attributive expression, and it is not usually counted as a

referring expression.

See also: deictic category; meaning; sense

Further reading: Hofmann 1993; Hurford and Heasley 1983; Saeed 1997.

RELEVANCE

A major determinant of meaningfulness. In pragmatics, the British

philosopher Paul Grice developed his notion of conversationalimplicature based on the maxims of quality, quantity, relevance and

manner. The British linguist Deirdre Wilson and the French philo-

sopher Dan Sperber developed relevance into a ‘super-maxim’ in

their 1980s relevance theory. This holds that the interpretation of an

utterance depends on resolving the context of the utterance to

produce the optimum level of information with the least processing

effort. Relevance theory produces a scale of interpretations of an

utterance from strong relevance to weak relevance. The theory isuseful for dealing with non-asserted meanings, where interlocutors

make inferences about each other’s intentions. The assumption of

relevance in communicativeness can explain the implicatures con-

veyed when all the Gricean maxims are flouted.

See also: conversational implicature; cooperative principle; pragmatics

Further reading: MacKenzie 2002; Sperber and Wilson 1995.

RHETORIC

Perhaps the oldest branch of linguistic study, the term rhetoric has

altered its meaning over its 3,000-year old history, but at heart

RELEVANCE

246

refers to the study of language and its effects. Originally part of the

classical Greek education ‘trivium’ with grammar and dialectic(reasoning), rhetoric was concerned with the art of persuasion and

oratorical authority. Aristotle divided rhetoric into ethos (the statusand authority of the speaker), pathos (his empathetic and emotional

disposition) and logos (rationality and logical organization of lan-

guage). Later Roman and Renaissance revisions established the five

‘canons’ of rhetoric as inventio (invention), memoria (memory),

pronuntiatio (delivery), dispositio (the organization of ideas) and

elocutio (style and register). The last of these transforms into sty-

listics in the modern era.

Rhetoric has again become popular as an umbrella term for asocially situated and ideologically aware form of language study. In

the USA, it has particularly been used to refer to college composi-

tion courses. Elsewhere in the world, it has been used in connection

with creative-writing courses which aspire to something more rig-

orous than impressionistic mutual congratulation.

See also: stylistics

Further reading: Sloane 2001.

ROOT

The minimal common form, or morpheme, which appears in all the

different forms of a single word. In Spanish, the root of the verb

meaning ‘sing’ is cant-, and this appears in all forms of the verb,such as canto ‘I sing’, cantabamos ‘we were singing’, canto ‘s/he

sang’, cantaras ‘you will sing’, cantarıan ‘they would sing’, cantemos

‘we might sing’, cantando ‘singing’, and so on. Likewise, in Arabic,

the root of the verb meaning ‘write’ is ktb, which appears in all

forms of the verb, including derivatives: katab ‘he wrote’, yiktib ‘he

will write’, maktuub ‘written’, kaatib ‘writer’, kitaab ‘book’, kutub

‘books’, and so on.

A root must be carefully distinguished from a stem, which con-sists of a root plus some additional material. In English, though,

this difference is rarely significant.

See also: morphology; stem

Further reading: Jackson 2002; Katamba and Stonham 2006.

ROOT

247

RULE

A statement expressing a generalization about the facts of a lan-

guage. Since ancient times, descriptions of languages have featuredrules of one sort or another, but, before the rise of modern linguis-

tics, these rules were usually expressed very informally, in ordinary

language. Linguists, in contrast, have usually been at pains to state

their rules with maximal explicitness; to this end, they have devised

various notational conventions, which allow rules to be stated in a

manner that is at once fully explicit, maximally economical and

embedded within a particular theoretical framework. This drive

towards explicitness was promoted by the American structuralistsduring the 1940s, but these linguists were more interested in dis-

tribution, and rules only became fully prominent with the rise of

generative grammar in the 1950s.

Rules can be stated at every level of linguistic description, from

phonology to pragmatics. A rule need not be exceptionless. For

example, the rule of English that says that a noun forms its plural

by adding <-s> is genuine, and it applies to most new nouns

entering the language, but there are some well-known historical orborrowed exceptions, like feet, children and radii.

It is important to realize that rules in linguistics are statements

about actual linguistic behaviour. Earlier grammarians, with their

devotion to prescriptivism, often stated ‘rules’ which were not rules

at all, but only their opinions about what should be considered

good usage, such as ‘Don’t end a sentence with a preposition’.

Linguistically unsophisticated people often assume even today that

this is what rules are, but such a conception has no place in seriouslinguistic work, which is embedded firmly within descriptivism.

See also: descriptivism; notational convention; prescriptivism

Further reading: Sampson 1975.

SAUSSUREAN PARADOX

How can speakers continue to use a language effectively when that

language is constantly changing? The Swiss linguist Ferdinand de

Saussure was the first to demonstrate that a language is not just a

collection of linguistic objects like speech sounds and words;

instead, it is a highly structured system in which each element is

RULE

248

largely defined by the way it is related to other elements. This

structuralist view of language has dominated linguistic thinking ever

since. But it immediately leads to a puzzle. We know that every

language is constantly changing. So: how can a language continue tobe a structured system of speech sounds, words, grammatical forms

and sentence structures when all of these are, at any given moment,

in the middle of any number of changes currently in progress?

This paradox greatly puzzled linguists for generations. Today,

though, we are well on the way to resolving it at last. The key

insight has come from the study of variation in language. Though

variation was formerly dismissed as peripheral and insignificant, we

now realize that variation in fact forms a large part of the verystructure of any language, that speakers make use of that variation

just as they make use of other aspects of language structure, and

that variation is the vehicle of change, as speakers simply shift the

frequencies of competing variant forms over time.

See also: language change; quantitative approach; structuralism; variation

Further reading: Milroy 1992; Trask 1996; Weinreich et al. 1968.

SCHEMA

A set of knowledge that is idealized from experience. Though the

notion of idealized mental knowledge appears in the nineteenth

century, it was developed in the 1970s specifically in relation to

language problems in artificial intelligence research. It becameapparent that machine language needed more than a semantic and

even pragmatic module in order to be able to pass as natural

human language. Giving the computer a restricted ‘experience’ would

provide the rich knowledge set that humans use to ‘fill out’ the

underdeterminacy of language. The classic example is the ‘restaurant

schema’, which explains how people can negotiate and understand

objects and utterances even in restaurants that they have never been

in previously, on the basis of their idealized mental respresentation.Various problems with schema theory have led to its adaptation

in a range of world-based approaches. For example, a schema is not

a rich world so it is not clear how far inferencing is valid; it is not

really a social world, so it does not explain how interlocutors can

communicate with a shared schema; the boundaries of a schema are

not well defined, so a ‘restaurant schema’ might be an ‘eating out

SCHEMA

249

schema’, or it might be a general collection of cafe schema, pub

schema, dinner-party schema, and so on. The ambiguities in schema

theory have, however, led to it being used in literary settings in

stylistics.

See also: cognitive linguistics; frame; possible worlds theory; stylistics;text world theory

Further reading: Cook 1994; Schank 1982, 1986; Schank and andAbelson 1977; Stockwell 2003.

SEGMENT

Any one of the discrete units which occur in sequence in speech. In

phonetics, a segment is a speech sound; in phonology, it is a phoneme.

But both levels of analysis recognize that a single word or a longer

piece of speech consists of a sequence of discrete units, occurring

one after the other. The word pat, for example, may be represented

at the phonetic level as [phæt], and at the phonological level as /pæt/,with a sequence of three segments in either case.

All those aspects of phonology which pertain to segments and their

behaviour are collectively known as segmental phonology, in contrast

to the study of suprasegmental phenomena, which must be described

with reference to phonological units larger than a single segment.

As a general rule, a given piece of speech will contain the same

number of segments at both levels, but there are exceptions. For

example, in many American accents, the word can’t contains foursegments at the phonemic level, /kænt/, but only three at the pho-

netic level, [khæt]. Here the phonemic segment /n/ loses its seg-

mental nature at the phonetic level and is realized only as a feature

of nasalization occurring on the vowel [æ].

See also: phoneme; speech sound; suprasegmental

Further reading: Ball and Rahilly 1999; Catford 1977.

SELECTION RESTRICTION

A restriction on the combining of words in a sentence resulting

from their meanings. It is easy to construct apparently grammatical

SEGMENT

250

sentences which are senseless because the meanings of the words in

those sentences cannot be combined in a comprehensible way: You

have deceived my watermelon; The square root of seven is green and

squishy; She dropped her shyness into the pond with a splash. Therestrictions which are violated here are selection restrictions, or

selectional restrictions. For example, the verb deceived here can be

regarded as requiring the selectional component [+human] as a

direct object, and watermelon is [–human].

In the early 1960s, there was an attempt by Noam Chomsky to

treat such restrictions as part of the grammar, and hence to regard

my examples as ungrammatical, but this idea was quickly dropped

as unworkable, and today such restrictions are universally regardedas belonging to semantics or (more usually) pragmatics. In the

semantic view, we regard my examples as impossible to interpret at

all; in the pragmatic view, we consider merely that it is difficult or

impossible to find a context in which one of these would be a

plausible thing to say.

See also: meaning; well-formedness

Further reading: Allan 1986; Leech 1974; Radford 1988.

SEMANTIC ROLE

Any one of several ways in which a person or thing may be involved

in an action or a state of affairs. The notion of semantic roles (alsocalled participant roles or deep cases or thematic roles or theta roles)is important in many approaches to linguistic description, particu-

larly in those approaches which embrace functionalism. The idea is

that a given entity which is involved in some event must play some

identifiable part in that event. For example, in the sentence Susie

tightened the nut with a spanner, Susie is an Agent (she is the insti-

gator of the action), the nut is a Patient (something is happening to

it), and a spanner is an Instrument (it is being used to accomplish

some purpose). In contrast, in Susie received a letter, Susie is aRecipient (something is arriving at her), while a letter is a Theme

(nothing is happening to it except that it is being moved).

Analysis of sentences or texts in terms of semantic roles may be

illuminating, but analysts often do not agree as to which semantic

roles should be recognized, and it is frequently very difficult to

assign roles in a principled manner. For example, in Susie filled the

SEMANTIC ROLE

251

bucket with water, it is clear that Susie is an Agent, but what are the

bucket and water? Location? Goal? Patient? Theme? Instrument?

Because of the difficulty of answering such questions, many lin-

guists have preferred to reject semantic roles altogether in con-structing their descriptions, but many others are convinced that

semantic roles are of fundamental importance in spite of the

difficulties.

See also: semantics

Further reading: Frawley 1992; Halliday 2004; Hurford and Heasley1983; Palmer 1994; Radford 1988; Saeed 1997; Thompson 2004.

SEMANTICS

The branch of linguistics which studies meaning. The study of

meaning has something of a chequered history in linguistics. People

have since ancient times been interested in questions of meaning,

but very little progress was made before the late nineteenth century,and semantics did not really exist as a distinct field. Around that

time, however, the French linguist Michel Breal, who coined the

term semantics, made a serious and largely successful attempt to

introduce semantics into European linguistic work. And, once the

Swiss linguist Ferdinand de Saussure had made the linguistic sign

the cornerstone of his influential theories, semantics was here to

stay in European linguistics.

Oddly, some of the most important work in semantics was beingdone from the late nineteenth century onwards by philosophers, but

it was a long time before linguists became aware of this work and

began to join forces with the philosophers.

American linguists were comparatively reluctant to consider

semantic questions. The two principal figures in the early twentieth

century, Edward Sapir and Leonard Bloomfield, did not entirely

neglect the matter, but they had little to say about it, and Bloom-

field’s successors, the American structuralists of the 1940s and1950s, were so pessimistic about the chances of applying linguistic

techniques successfully to the seeming swamp of meaning that they

effectively defined linguistics as a field excluding semantics.

In the 1960s, however, shortly after the Chomskyan revolution, a

few American linguists began to be interested in semantic ques-

tions. They were ignorant of the vast philosophical literature, and

SEMANTICS

252

their first attempts were fumbling, but before long they had caught

up and were beginning to join forces with the philosophers. The work

of the American philosopher Richard Montague had an enormous

effect upon the linguists, and Montague’s ideas have since becomethe basis of a great deal of important work in linguistic semantics.

One of the most important advances was made gradually during

the 1960s, when it was realized that there were two fundamentally

different types of linguistic meaning. One type of meaning is

intrinsic to the linguistic form containing it, and is always present

in that form, while the second type of meaning results from the

interaction between the linguistic form of an utterance and the

context in which it is uttered. Today we understand semantics asproperly the study of the first type, while to the study of the second

type we give a new name: pragmatics (a term actually coined by the

American philosopher C.S. Peirce in the nineteenth century). Fail-

ure to make this distinction had earlier proved a severe obstacle to

progress.

In recent years approaches to semantics have proliferated, and

the subject is now one of the liveliest areas in linguistics. The

majority of the current approaches represent versions of formalsemantics: attempts at elucidating meaning by developing particular

versions of formal logic that can capture aspects of meaning.

Among the most influential threads of investigation have been

truth-conditional semantics, which attempts to reduce meaning to

questions of truth and falsehood, model-theoretic semantics, whichoperates in terms of miniature artificial universes called models, andsituation semantics, which embeds the study of meaning within

miniature contexts called situations.Most of the activity just mentioned is concerned with the

semantic interpretation of sentences, but the study of word mean-

ing, called lexical semantics, has also been extensively developed.

See also: meaning; pragmatics; selection restriction; sense relation

Further reading: Cruse 1986; Frawley 1992; Hofmann 1993; Hudson1995; Hurford and Heasley 1983; Kreidler 1998; Saeed 1997.

SEMIOTICS

The study of the social production of meaning from sign systems.

Semiotics, also called semiology, traces its origins to the work of the

SEMIOTICS

253

Swiss linguist Ferdinand de Saussure in the early twentieth century,

and particularly to Saussure’s idea of the linguistic sign. Never-

theless, semiotics has never been regarded as part of linguistics, and

it has been developed almost exclusively by non-linguists, particu-larly in France, where it is often considered an important discipline.

In the English-speaking world, it enjoys almost no institutional

recognition.

Though language itself is taken to be the paradigm case of a sign

system, in practice most semiotic work has concentrated upon the

analysis of such varied domains as advertising, cinema and myths.

The influence of the central linguistic concept of structuralism

(another of Saussure’s contributions) has led semioticists to attemptstructuralist interpretations of a wide range of phenomena. An

object of study, such as a film or a cycle of myths, is viewed as a

text which communicates meaning, and that meaning is assumed to

derive from the orderly interaction of meaning-bearing elements, the

signs, which themselves are embedded in a structured system, some-

what analogously to the meaning-bearing elements in a language.

In spite of its deliberate emphasis upon the social nature of the

sign systems examined, semiotics tends to be highly abstract and attimes seemingly impenetrable. In recent years, however, semioticists

have increasingly turned to the study of popular culture, and

semiotic treatments of soap operas and pop music are now com-

monplace. Under the new guise of social semiotics, the field has

joined forces with critical discourse analysis and cultural studies to

investigate the interactions between visual and verbal in the modern

world.

See also: linguistic sign; symbolic system

Further reading: Cobley 2001; Eco 1976; Leeds-Hurwitz 1993; Sebeok1984, 1994; Tobin 1990.

SENSE

The central meaning of a linguistic form, regarded from the point

of view of the way it relates to other linguistic items. The central

meaning of a linguistic form, such as cat, can be approached from

at least two different points of view. One way is to consider all the

things in the world and decide to which ones the form cat can rea-

sonably be applied; this approach leads to what we call the denotation

SENSE

254

of cat. The other is to compare the meaning of cat with the mean-

ings of other linguistic forms, such as lion and dog, and to decide

what semantic characteristics the form cat has which allow it to be

applied to some things but not to others. This leads to what we callthe sense of the form.

A very crude interpretation of the sense of cat might look some-

thing like this: carnivorous mammal, has four legs and a long tail,

has fangs and sharp retractable claws, has excellent eyesight even in

dim light, . . . You can easily see some shortcomings here: Manx

cats have no tail, cheetahs cannot retract their claws, and so on.

What we are describing here is really a stereotype of a cat, a maxi-

mally typical (but hypothetical) cat, and a real animal qualifies as acat if it matches the stereotype sufficiently well, even if not per-

fectly. (Note: a stereotype is different from a prototype. A prototype

of a cat would be a single real cat which was regarded as so emi-

nently typical of cats generally that we might like to hold it up as a

model of perfect cathood, and cats encountered in real life would

be determined as good or bad examples of catness on that basis.

But note that not everyone uses the terms stereotype and prototype

in exactly the manner described here.)In formal versions of semantics, the sense of a linguistic form is

often formalized as its intension – that is, as the set (in the formal

mathematical sense) of all the properties which an object must have

before the form can be properly applied to it.

Sense is most commonly contrasted with reference, but it is

essentially just a different way of looking at the same kind of

meaning singled out as denotation in a different approach.

See also: denotation; prototype; reference; stereotype

Further reading: Allan 1986; Hurford and Heasley 1983; Ungerer andSchmid 1996.

SENSE RELATION

Any of several ways in which the meanings of words may be rela-

ted. Words do not have meanings in isolation; instead, the meaning

of a word is usually related in important ways to the meanings of

other words. Some of the most prominent of these relations in

meaning are known collectively as sense relations, and there are

several kinds.

SENSE RELATION

255

In synonymy, two words have identical or nearly identical mean-

ings: cat and feline, pail and bucket, violin and fiddle, fruitful and

productive. Such pairs of synonyms are not always completely

interchangeable: the words may differ in degree of formality, as withfelon and crook, or they may differ in their connotations (associa-

tions), as with rabbit and bunny. However, if both words can be

applied in principle to exactly the same things, they are cognitivelysynonymous.

In antonymy, two words have opposite meanings. Some pairs of

antonyms are gradable antonyms, representing extremes along a

continuum, as with hot and cold or big and small. Other pairs are

binary antonyms, which are mutually exclusive and which betweenthem exhaust the possibilities, as with alive and dead.

Somewhat different are converse pairs, such as wife and husband

or above and below: if I am your husband, then you are my wife; if

the table is below the clock, then the clock is above the table.

In meronymy, one word denotes a part of another. For example,

hand is a meronym of arm, and both words are meronyms of body.

In hyponymy, one word denotes a special case of what is denoted

by the other. For example, spaniel is a hyponym of dog, which inturn is a hyponym of animal, while dog is a superordinate of spaniel.In plesionymy, a word which is a rough approximation or specifica-

tion of another word can be stated along with it: it’s not raining,

more drizzling; my father is a shopkeeper, a greengrocer.

Words which relate semantically in appropriate ways (little girl,

children play football) are philonyms; words which clash semanti-

cally (the trees looked on; mean streets) are xenonyms.In all cases, rhetorical effects are generated where one item is

selected in place of another similar sense relation: insult (That

woman is a mammal), metaphor (I poured myself home), formality

(He composed an epistle), euphemism (Washington ordered the

bombing), and so on.

See also: meaning; semantics

Further reading: Cruse 1986; Hurford and Heasley 1983.

SENTENCE

The largest purely grammatical unit in a language. Of course, there

are larger linguistic units than sentences: individual paragraphs (in

SENTENCE

256

writing), individual turns (in conversation) and individual dis-

courses. But these larger units are no more than very weakly linked

by purely grammatical means. The largest linguistic unit which is

held together by rigid grammatical rules is the sentence.We need to clarify this term a little, since it is frequently mis-

understood. For most linguists, in most circumstances, a sentence is

an abstract linguistic object: specifically, it is a linguistic object put

together entirely in accordance with the rules for constructing sen-

tences in a language, rules which have to be identified (in a linguis-

tic description) by patient and painstaking investigation. More

particularly, a sentence does not have to be something which

somebody might reasonably say, and not everything that we mightreasonably say is a sentence.

Consider this exchange. Mike: Where’s Susie? Alice: In the library.

Here Mike’s utterance represents a sentence, but Alice’s response

does not: even though it is a perfectly normal and unremarkable

thing to say, *In the library is not a sentence of English (the asterisk

indicates this fact), because it is not constructed according to the

rules for making English sentences. Instead, it is only a fragment ofa sentence: we do not always speak in complete sentences, and wevery often use fragments like this one.

Now consider Noam Chomsky’s famous example sentence: Col-

ourless green ideas sleep furiously. Chomsky’s point is that, even

though this thing makes no sense at all, it is constructed in accor-

dance with all the rules for making sentences in English, and hence

it is a grammatical (well-formed) sentence of English.

A further point is that a sentence is not just a string of

words; rather, it is a string of words with a grammatical (syntactic)structure assigned to it. Consequently, an ambiguous string of

words like Visiting relatives can be a nuisance or I saw her duck

represents two (or more) quite different sentences, each with its

own structure. We often take advantage of this possibility of

assigning different structures to identical or similar strings of words

for humorous effect, as in the old gag Time flies like an arrow; fruit

flies like a banana and in the punchline of a certain cat-food com-

mercial: Cats like Felix like Felix. Pragmatics makes the useful dis-tinction between the abstract sentence and the actual spoken or

written utterance.

See also: ambiguity; clause; syntax; utterance; well-formedness

Further reading: Brown and Miller 1991; Lyons 1968; Matthews 1981.

SENTENCE

257

SEX DIFFERENCES IN LANGUAGE

Differences between the speech of men and women. In some lan-

guages, there are very conspicuous differences between men’s andwomen’s speech: men and women may use different words for the

same thing, they may use different grammatical endings, they may

even use different sets of consonants and vowels in their pronunciation.

English has nothing quite so dramatic as this, but several decades

of research have turned up some interesting differences even in

English – though not all of the early claims have been substantiated

by later work. For example, it has been suggested that women use

more tag questions than men – as in It’s nice, isn’t it? – as if to seekapproval for their opinions, but this has not been borne out by

investigation. It has also been suggested that men swear more than

women, but this too appears not to be so, at least among younger

speakers.

On the other hand, it does appear to be true that certain words

are more typical of women, including terms of approval like cute,

divine and adorable and specific colour terms like beige, burgundy

and ecru. Women have also been observed providing more back-channel (hmm, yeah, ah-ha), asking more questions, inviting

approval and agreement, and generally being more socially sup-

portive in conversation than men. The British sociolinguist Jennifer

Coates and the American sociolinguist Deborah Tannen have dis-

covered that men and women organize their conversations very

differently. Men tend to be rather competitive; women engage in

highly cooperative conversations. Men engage in floor-holding;

women latch on to topics, completing the speaker’s turn, or inter-ject approval or disagreement. That is, a conversation among

women is a collaborative enterprise, with all the women pulling

together to construct a satisfactory discourse which is the product

of all of them, while a conversation among men is rather a

sequence of individual efforts.

These differences can lead to serious misunderstanding in mixed

conversations. While her husband or boyfriend is speaking, a

woman may constantly contribute supporting remarks in thenormal female fashion, but the man may well interpret these

remarks as interruptions (which they are not) and become very

annoyed. In fact, it is quite clear that, in mixed conversations, men

interrupt women far more than the reverse.

It is clear that such differences in behaviour are indexes not

directly of gender but of power. Men in subservient job situations

SEX DIFFERENCES IN LANGUAGE

258

and women in powerful settings have been observed using the power-

marked pattern of discourse, regardless of gender.

See also: gender; power; turn-taking

Further reading: Cameron and Kulick 2003; Coates 1996, 1998, 2004;Connell 1995; Holmes 1992; Holmes and Meyerhoff 2003; Romaine1994; Tannen 1991.

SEXIST LANGUAGE

Language which, deliberately or unconsciously, is patronizing or

contemptuous towards one sex, usually women. Sexism, of course,

is not specifically a linguistic issue, but it shows up in languages in

various ways, some of them rather deeply embedded. Almost with-

out exception, sexist usages are patronizing of women. Here are a

few examples.

As is well known, English has only the sex-marked singular pro-

nouns he and she, and hence a speaker addressing or talking abouta mixed group has a problem: Somebody has forgotten his umbrella

is sexist, while Somebody has forgotten his or her umbrella is almost

unbearably clumsy. In this case, popular speech usually solves the

problem by using their: Somebody has forgotten their umbrella. But

some people find this distasteful, and it doesn’t really work very

well in cases like Any student who considers themself adequately

prepared is requested to present themself for their oral examination.

Many pairs of sex-marked words have developed very differently.A master is a powerful or skilful man; a mistress is a woman kept

for sexual purposes. A courtier is a polished man of high social

status; a courtesan is just an upmarket whore. There is nothing

wrong with calling a man a bachelor, but calling a woman a spinster

is contemptuous. Even a single word may behave differently: in

American English, at least, when you call a man a pro, you mean

that he is experienced, competent and reliable; when you call a

woman a pro, you mean she’s a prostitute.The female suffixes -ess, -ette, -ienne, -ine, -ix, and -euse cause

particular problems: consider the sense of diminutiveness or secon-

dariness in actress, poetess, jewess, authoress, sculptress, stewardess,

waitress, governess, comedienne, heroine, masseuse, usherette, brunette,

aviatrix and the marked forms for confidante, blonde, divorcee and

starlet. But there are many other such cases. Men play golf and tennis,

SEXIST LANGUAGE

259

while women play women’s golf and (worse) ladies’ tennis. Many

people still qualify professional names with lady doctor, female pro-

fessor or male nurse. Among the most blatant examples of all are the

following, both genuine: The assailant attacked his next-door neigh-

bour’s wife (the woman was not his neighbour?); The pioneers trekked

across the prairies with their cattle, their seed-corn and their wives

(the wives were only there to cook, clean, sew and raise the children

while their husbands were busy pioneering?). Many of these forms

are beginning to pass into archaism, but they can still be found.

Once rarely remarked on, sexist language has been drawing the

fire of feminists for several decades now, and a number of linguists

have turned their attention to the issue. Attempts at stamping outsexist usages have enjoyed some success, and terms like fireman,

postman and chairman are now commonly replaced by firefighter,

letter carrier and chairperson (or simply chair); similarly, the use of

man or men to denote human beings in general is slowly giving way

to human beings or humans. Here is a rare example of prescriptivism

being used for a socially useful purpose. However, other prescrip-

tions such as mistress copy (for the master copy of a document), and

attempts to avoid manhole and man-eating tiger have been morequestionable and less successful.

See also: language planning

Further reading: Cameron and Kulick 2003; Coates 2004; Hofstadter 1985.

SIGN LANGUAGE

A language whose medium is signs made with the hands and head.

Speech is the most familiar medium of language, but it is not the

only one possible. Deaf people cannot hear speech, and many deaf

people learn and use a sign language as their primary language,

often as their first language. Many different sign languages exist,

including British Sign Language (BSL) in the UK and AmericanSign Language (ASL) in the USA; these two are not related to eachother, and neither is related in any way to English.

A true sign language is not a crude approximation to a spoken

language; it is a genuine natural language with a large vocabulary

and a rich and complex grammar, and it is every bit as flexible and

expressive as a spoken language. The basic units are signs made

chiefly with the hands and the head; in ASL and BSL, these signs

SIGN LANGUAGE

260

can be modified in various ways to express shades of meaning like

‘many’, ‘often’, ‘slightly’, ‘quickly’, ‘very’ and ‘repeatedly’, notions

which would require separate words in English. An example: in

ASL, touching the middle finger to the forehead means ‘be sick’; bymodifying the distance, direction, and speed of movement, and the

number of touches, the signer can explicitly express (at least) any of

‘get sick’, ‘get sick easily’, ‘often sick’, ‘never stops being sick’,

‘sickly’, ‘slightly sick’, ‘very sick’, and ‘sick for a long time’. Indeed,

it has been remarked that the grammars of ASL and BSL are more

similar to the grammars of certain Native American languages,

such as Hopi, than to the grammar of English.

Just like creoles, sign languages emerge spontaneously wheneverdeaf children are brought together. A recent example is Nicaraguan

Sign Language, created and used by the deaf children whowere brought

together to receive an education after the Nicaraguan revolution of

1979. ASL and BSL, in fact, were deliberately invented by hearing

people in the nineteenth century, but the modern versions are vastly

more elaborate than the original ones, as native signers have

expanded the vocabulary and elaborated the grammar to meet their

needs. And, just like spoken languages, sign languages change overtime: already ASL is showing evidence of developing regional dialects.

So strong is the language instinct that a deaf child will eagerly

begin babbling with its hands, and it will seize upon any gestures it

sees and do its best to turn them into a sign language, even without

reinforcement; such a system is called home sign.Linguists were at first slow to appreciate that sign languages were

true languages, but, since the pioneering work of the American linguist

Ursula Bellugi in the 1970s, sign languages have come to be treatedon a par with spoken languages, though most introductory text-

books of linguistics provide regrettably little coverage of the topic.

See also: language; language instinct; natural-language processing

Further reading: Deuchar 1984; Klima and Bellugi 1979; Miles 1988;Morgan and Woll 2002; Padden 1988; Smith 1990; Steinberg 1993;Valli and Lucas 2001.

SLANG

Informal and often ephemeral linguistic forms. We all use our lan-

guage in different ways, depending on the circumstances. Most

SLANG

261

obviously, we speak differently in formal contexts and in informal

contexts. Especially when speaking informally, we often take pleasure

in resorting to slang: informal but colourful words and expressions.

Slang expressions are usually introduced by the members of aparticular social group; they may remain the property of that group

and serve as a badge of group identity, or they may instead become

much more widely known and used. The majority of slang forms

are transient: they are used for a few months or a few years, and

then they pass out of use, to be replaced by even newer slang terms.

Just in the past fifty years in English, something really excellent

has been described at times, by certain groups of speakers at least,

as groovy, fab, brill, tremendous, wicked, ace, spiffing, smart, out-

standing, cool, far out, awesome, sweet, triff, def, and countless other

terms. One or two of these are perhaps still in use, but the rest are

already quaint anachronisms. We have variously said of a man who

is drunk that he is loaded, soused, fried, pickled, sozzled, pissed,

blitzed, bombed, smashed or tired and emotional, or that he has had

a skinful or is three sheets to the wind. Slang terms tend to be over-lexicalized areas of life which are or have been taboo: sex, death,

excrement, drunkenness and intoxication, racism, sexism, homo-phobia and so on.

An unusual case of a long-lived slang term is booze for alcoholic

drinks; this has been in the language for centuries, but it is still

used, and it is still regarded as slang. On occasion, a slang term

may lose its slang status altogether; this has happened with mob, an

old slang word (short for mobile) which is now unquestionably

standard English. Slang is often playful and creative, and its pre-

sence is evidence of the vitality of a language.

See also: prescriptivism; vernacular

Further reading: Crystal 1995; Partridge 1970, 2002.

SOCIAL HISTORY OF LANGUAGE

The study of the history of language as a social institution. Tradi-

tionally, historical linguistics, and in particular the study of language

change, has focused upon internal structural changes: changes in

pronunciation, grammar and vocabulary. In the 1960s, however,

linguists began to turn their attention to the social context of language

change, to the ways in which changes are introduced and propagated

SOCIAL HISTORY OF LANGUAGE

262

by speakers, and to the social forces accompanying these changes,

and this type of investigation has proved very illuminating.

More recently still, a few scholars have begun to turn their

attention to broader issues in the history of language. Among theseissues are the choice between languages or language varieties in

particular communities at particular times, the social pressures

associated with these choices, the rules governing conversation

among various social groups in the past, the connections between

language and both individual and national identity, and the reasons

for changes in all these things over time. To this new discipline we

give the name social history of language.The social history of language is perhaps more obviously a

branch of social history than of linguistics, but it complements the

more familiar concerns of linguists by providing an overview of the

circumstances in which particular languages were used in particular

communities at various times in the past. Socially situated and

ideological explanations for language change are often persuasive.

See also: language planning

Further reading: Burke 1993; Crowley 1989, 1996; Crowley 1997; Fen-nell 2000; Honey 1997; Leith 1997; Lodge 1993; McCrum et al. 1992.

SOCIAL NETWORK

A means of measuring the degree of linguistic influence exerted at a

local level on and by individuals. A social network can be ‘scored’on the basis of how many people are intimately connected within it

(its density) and on the extent to which the connections between

people are not one-dimensional – so, for example, people are rela-

ted as friends, neighbours, colleagues, football team-mates, church-

goers, and so on (the multiplexity of the network). In a socio-

linguistic study in Belfast, Northern Ireland, the linguist Lesley

Milroy discovered that a dense and multiplex social network oper-

ated as a strong linguistic norm reinforcement mechanism. Accentfeatures (and innovations) within strong social networks were

maintained and diffused through the wider community.

See also: language change; prestige; sociolinguistics; speech community

Further reading: Milroy 1987a; Llamas et al. 2007.

SOCIAL NETWORK

263

SOCIAL STRATIFICATION OF LANGUAGE

Marked differences in the speech of people belonging to different

social classes in a community. In most communities of any size,there are conspicuous differences in the social status of people, and

these social classes are usually hierarchically ordered in overt pres-

tige, from highest to lowest.

It has long been realized that members of different social classes

tend to speak differently, but it is only since the 1960s that socio-

linguists, such as the American William Labov and the Briton Peter

Trudgill, have begun to undertake systematic studies of these dif-

ferences. Such studies often reveal a great deal of information aboutthe differing linguistic behaviour of the several social classes. For

example, in London, t-glottaling (the pronunciation of the /t/ in

words like water as a glottal stop) is maximally prominent amongst

lower working-class speakers, less prominent among middle-working-

class speakers, still less prominent among upper-working-class

speakers, and virtually absent amongst middle-class speakers of all

types. On the other hand, the respectively construction, as in Esther

and Larry drank whisky and brandy, respectively, is maximally pro-minent among higher-class speakers but absent altogether in working-

class speech.

These are examples of the social stratification of language: a

steady rise or fall in the frequency of particular linguistic forms as

we move through the social classes. And note that word ‘frequency’:

only sometimes is a particular form either categorically present or

totally absent in the speech of a particular class. Very often, it is

present in the speech of all or nearly all classes, but it differs infrequency, and ‘correct’ behaviour for a member of a class involves

getting the frequency right: using a form too frequently or too

rarely will make an individual’s speech sound anomalous to the

other members of his or her class.

See also: sociolinguistics; variation

Further reading: Labov 1972; Trudgill 1995; Llamas et al. 2007.

SOCIOLINGUISTICS

The branch of linguistics which studies the relation between lan-

guage and society. Though the social aspect of language attracted

SOCIAL STRATIFICATION OF LANGUAGE

264

early attention, notably from the great Swiss linguist Ferdinand de

Saussure at the beginning of the twentieth century, it was perhaps

only in the 1950s that serious investigation began. Pioneers like

Uriel Weinreich, Charles Ferguson and Joshua Fishman drewattention to a range of fascinating phenomena, such as diglossia

and the effects of language contact. But the key figure here is argu-

ably the American William Labov, who in the 1960s began a series

of investigations of variation in language, investigations which have

revolutionized our understanding of how speakers use their lan-

guages and which have finally resolved the Saussurean paradox.

Sociolinguistics may be usefully defined as the study of variation

in language, or more precisely of variation within speech commu-nities, since the purely geographical aspects of variation had been

studies for generations by the students of dialect geography and

dialectology, the study of regional dialects. In a speech community

of any size, there is considerable variation among individuals:

stockbrokers do not speak like plumbers, women do not speak like

men, young people do not speak like old people, and so on. More-

over, even a single individual is not confined to a single variety of

the language: you do not use the language in the same way whenyou are chatting to friends in a bar, when you are being interviewed

for a job, when you are writing an essay, and when you are being

introduced to royalty. Earlier linguists had, of course, noticed this

variation, but they were inclined to dismiss it as peripheral, as

inconsequential, even as a nuisance getting in the way of good

descriptions. Today, however, we recognize that variation is an

integral and essential part of language, and that absence of varia-

tion is almost pathological.Sociolinguistics in the USA is strongly focused on language var-

iation and change; elsewhere in the world, other matters that shade

into pragmatics and the sociology of language are also regarded as

part of the discipline.

See also: sex differences in language; social stratification of language;variation

Further reading: Bratt-Paulston and Tucker 2003; Cameron and Kulick2003; Chambers 2003; Chambers and Trudgill 1998; Chambers et al.2002; Coulmas 2005; Eckert and Rickford 2001; Fennell 2000; Harris2006; Holmes 1992; Hudson 1996; Llamas et al. 2007; Milroy andGordon 2003; Romaine 1994; Stockwell 2002a; Trudgill 1995; Ward-haugh 2005.

SOCIOLINGUISTICS

265

SOUND SYMBOLISM

An attempt at constructing a word whose sound directly conveys

(some aspect of) its meaning. The most familiar type of soundsymbolism is onomatopoeia, in which the meaning of a word is a

real-world sound and the form of the word attempts to mimic the

sound. Examples: bang, boom, murmur, hiss, quack, meow, clink,

ding-dong, thud. But other types exist.

English has a group of words with initial /sl-/, all with meanings

in the general area of ‘slippery, slimy, gooey’: slurp, slip, slide, slink,

slush, slop, slosh, sludge, slurry, slime, slug (the creature), slaver, and

so on. The sequence /sl-/ in these cases is a phonaestheme, and thiskind of sound symbolism is phonaesthesia.

Yet another type of sound symbolism is represented by two

Basque words for ‘butterfly’, tximeleta and pinpilinpauxa (<tx> =

English <ch> , <x> = English <sh> ). Here the fluttery sound of

the words seems designed to mimic the fluttery appearance of the

creature.

One more rather frequent type is represented by the ideophonesof many languages, which attempt to mimic types of motion.Examples from the Carib language Apalai: seky seky ‘creep up’, ty

ty ty ‘person walking’, wywywywy ‘hammock swinging’, uroruro

‘trees falling’, tututututu ‘fast approach’, and so on.

Sound symbolism constitutes a partial exception to the more

usual arbitrariness of linguistic forms.

See also: arbitrariness; iconicity

Further reading: Crystal 1988, 1995, 1997; Palmer 1996.

SPEECH

Spoken language, either in general or in particular instances. The

term speech is used in linguistics in three rather different ways,

which sometimes need to be distinguished.First, speech is a medium for language; in this sense, the term

contrasts with writing and with sign language.

Second, speech is the overall linguistic behaviour of people who

speak, including any patterns visible in that behaviour. This is the

sense of the term in compound terms like speech community and

speech act.

SOUND SYMBOLISM

266

Third, speech is the real utterances produced by real people on

real occasions. This is the sense of the term in speech error, and it is

exactly equivalent to the terms parole and performance.

See also: medium; performance

Further reading: Biber 2006; Knowles 1987; Shockey 2003.

SPEECH ACT

An attempt at doing something purely by speaking. There are verymany things that we can do, or attempt to do, simply by speaking.

We can make a promise, ask a question, order or request somebody

to do something, make a threat, name a ship, pronounce somebody

husband and wife, and so on. Each one of these is a particular

speech act.In the majority of cases, it makes no sense to ask whether an

utterance constituting a speech act is true or false. Utterances like

Clean up your room!; Can you lend me a pen?; I promise to buy you a

teddy bear; and I dub thee knight, Sir Clarence have no truth value,

but they may be more or less appropriate to the circumstances, or,

as we say, more or less felicitous. An utterance like Clean up your

room! is infelicitous if I have no authority over you, and I now

pronounce you husband and wife fails unless a number of obvious

conditions are met. The conditions required for a speech act to be

successful are accordingly often called felicity conditions.Speech acts belong to the domain of pragmatics, and their

study, called speech-act theory, is a prominent part of that dis-

cipline. The study of speech acts was introduced by the British

philosopher J.L. Austin in the 1960s, and it has been developed by a

number of others, notably the philosopher John Searle. Austin ori-

ginally distinguished three aspects of a speech act: the locutionaryact (the act of saying something and its basic content), the illocu-tionary act (what you’re trying to do by speaking), and the perlo-cutionary act (the effect of what you say). Today, however, the termspeech act is often used to denote specifically an illocutionary act

(promising, threatening, informing, persuading, defending, blam-

ing, and so on), and the intended effect of a speech act is its

illocutionary force.

See also: performative; pragmatics

SPEECH ACT

267

Further reading: Grundy 1995; Levinson 1983; Mey 1993; Saeed 1997;Schiffrin 1994; Thomas 1995; Verschueren 1999; Yule 1996.

SPEECH COMMUNITY

A group of people who regularly interact by speaking. A speech

community may be large or small, and it may be highly homo-

geneous or decidedly heterogeneous. What matters is that every-

body in it should regularly speak to at least some of the other

people in it, and that the community should not be broken up

by sharp boundaries across which speaking rarely or never takesplace.

Not infrequently, an individual may belong to several speech

communities of different sizes. For example, a rural American in

the Midwest belongs to a small local community with whom he

interacts intensely, a much larger surrounding community with

whose members he interacts less frequently, and, less centrally, to

the entire community of English-speakers in the world. In the

majority of cases, though, when we speak of a speech community,we have a smaller community in mind, not a larger one.

It is not necessary for everybody in a community to speak the

same variety of a language. For example, London and New York

include English-speakers speaking a striking range of different

varieties of English. And it is not even necessary for everybody in

the community to speak the same language at all. For example,

Singapore has large numbers of native speakers of Malay, Canto-

nese, Tamil and English; few people can speak all four, but almosteverybody can speak at least two, and Singapore does not consist of

several isolated communities having little to do with one another.

See also: bilingualism; community of practice; diglossia; social network;social stratification of language

Further reading: Chambers 2003; Edwards 1994; Holmes 1992; Llamaset al. 2007.

SPEECH EVENT

A significant piece of speaking conducted according to rules. All

speech possesses structure of some kind, but there are certain pieces

SPEECH COMMUNITY

268

of speaking which are rather special. Each one has a recognizable

beginning and end and is constructed according to certain rules

known to both speakers and listeners. Examples include a uni-

versity lecture, a sermon, an after-dinner speech, a debate and ajob interview. Such a highly structured piece of speech is a speechevent.

A speech event involves participants who assume clearly defined

roles, and it takes place in a well-defined setting. The rules govern-ing the event are clearly defined and known to all participants;

violating these rules is a serious lapse. Naturally, linguists are often

interested in identifying the rules governing particular sorts of

speech events.(Note: a few people apply the term speech event more broadly, to

include ordinary conversations, brief exchanges and even speechsituations [such as cocktail parties], but these extended usages are

not usual in linguistics.)

See also: discourse

Further reading: Duranti 1997.

SPEECH SOUND

One of the individual sounds produced in sequence during speech.

When we speak, the several speech organs are all in constant

motion: the lips, the jaw, the velum, the glottis and the several parts

of the tongue are all moving about at their own pace, with only theoccasional moment of motionlessness. Nevertheless, we hear the

result as a linear sequence of individual sounds, each following the

last, and each having its own distinguishing characteristics. Each

one of these perceived sounds is a speech sound.For example, a typical pronunciation of the English word cleaned

can be represented in phonetic transcription, with an absolute

minimum of phonetic detail, as [klind]. Here the transcription

shows that we hear this word as a sequence of five speech sounds,namely [k], [l], [i], [n] and [d]. In purely phonetic terms, this is an

oversimplification: the voicelessness of the [k] is extended into the

[l], the nasality of the [n] begins during the [i], and there are many

other phonetic details not represented in this very simple

transcription – and yet what we hear is something reasonably

represented as [klind].

SPEECH SOUND

269

It is important to realize this perceptual nature of speech sounds.

For example, a typical American pronunciation of can’t differs from

cat only in that the first has a nasalized vowel: [kæt] versus [kæt].

Yet American speakers perceive the first as consisting of fourspeech sounds, [kænt], but the second as only three, [kæt]. Speakers

hear the speech sounds that are ‘supposed’ to be there, rather than

objective physical reality.

Speech sounds are classified into two types: consonants, which

involve a significant obstruction to the flow of air, and vowels,

which don’t.

See also: consonant; phoneme; segment; vowel

Further reading: Ashby 1995; Ball and Rahilly 1999; Ladefoged 1993;Ladefoged and Maddieson 1996.

SPEECH THERAPY

Also called language pathology, this is the application of linguisticknowledge to assist speakers with speech disabilities. These can

range from the major aphasic effects of brain trauma such as

strokes and physical features such as cleft palates, to temporary

disfluencies such as stammering (stuttering in the USA), lisping and

others. Many of these are features of child language acquisition,

and much progress can often be made without surgical or physical

intervention by the application of training techniques both for the

sufferer and parents or carers. For example, about 5 per cent ofchildren under five years of age experience some significant dis-

fluency, often related to periods of rapid language-acquisition pro-

gress. This can present itself as repetition of initial syllables,

repetition of single words – especially connectors (but, and, so) – or

‘blocks’ like an unreleased phonetic stop. Most children develop

fluency again without any intervention, but typical ways of treating

stammering include: paying attention to the child; slowing the

pace of life down, including speed of language; avoiding over-complicated syntax or vocabulary; and encouraging rather than

focusing on the problem.

See also: aphasia; applied linguistics

Further reading: Fabbro and Asher 1999.

SPEECH THERAPY

270

SPELLING

A conventional way of writing the words of a language using an

alphabetic writing system. Most languages are not normally writtenat all, and a few more, like Chinese and Japanese, are usually writ-

ten in non-alphabetic writing systems. The rest of the world’s lan-

guages are regularly written using some alphabet or other (there are

dozens in use), and, in each language, its conventional system for

representing particular words in writing is its spelling system.

In a particular type of ideal spelling system, called phonemicspelling (or, informally but inaccurately, ‘phonetic spelling’), each

single phoneme is always spelled in the same way, and differentphonemes are always spelled differently. Most spelling systems do

not approach this ideal, though a few, such as the Finnish and

Italian systems, come moderately close.

The reasons for departing from this ideal are numerous, impor-

tant and often intractable. The notoriously messy and irregular

spelling of English illustrates all of them.

First, once a spelling is established, it becomes part of ordinary

education, and speakers are reluctant to change it, even when thepronunciation of the language changes, as it frequently does. This is

the reason for such odd spellings as knee, night, of, write, steak,

iron, sugar and lamb: once they represented the pronunciation per-

fectly, but the pronunciations of these words have changed sub-

stantially, while the spellings have not been changed.

Second, it is convenient to use similar spellings for related words

even when those words are pronounced very differently. If we spel-

led photograph and photography just as we pronounce them, theobvious relation between them would no longer be visible.

Third, when words are taken from foreign languages, it seems

natural to people who know those languages to retain the foreign

spelling, even when this doesn’t match the English pronunciation.

This is the reason for spellings like machine, Zeitgeist, concerto,

lasagne and chic, and, with complications, for photograph, xylo-

phone and psychology.

Fourth, spellings are sometimes contaminated by the spellings ofother words. So, for example, isle is taken from Old French and

retains its Old French spelling, while the native English word island,

which was formerly spelled iland, has acquired its s by contamina-

tion (and so has aisle).

Fifth, and most importantly, we do not all pronounce our words

in the same way. So, for example, some people pronounce caught

SPELLING

271

and court identically, while others pronounce them very differently,

and so what would be a perfect spelling system for one group would

be hopeless for the other – and the same is true of caught and cot,

horse and hoarse, pull and pool, marry and merry (and also Mary),dew and do, poor and pour, father and farther, and countless other

such pairs. On top of this, many individual words are pronounced

in very different ways by different people: either, economics, tomato,

grass, suggest, library, sterile, contribute and many, many others.

There are, in fact, many advantages in having a spelling system that

does not privilege a single accent; this, and natural conservative

inertia, has defeated just about every attempt at prescriptivist spel-

ling reform ever proposed, at least for English.

See also: orthography; writing system

Further reading: Coulmas 1996; Crystal 1997; Robinson 1995; Sampson1985.

STANDARD LANGUAGE

That variety of a language considered by its speakers to be most

appropriate in formal and educational contexts. Consider the case

of English. English is spoken as a mother tongue by some 400

million people, as an everyday second language by tens of millions

more, and as a fluent foreign language by further millions. Many of

these people also write in English.

The English used by all these people is far from uniform: thereare both regional differences and differences among social groups

within a single community, and today sociolinguists often speak of

Englishes to denote this whole range of varieties. But one of these

varieties has a very special status: this is the variety called standardEnglish.

Standard English is the form of English acquired through edu-

cation; indeed, acquisition of standard English is a large part of

what we understand as education in the English-speaking world.Most broadcasting, and almost all publication, is couched in stan-

dard English.

Standard English may be spoken in any of a large range of

regional accents; no particular accent is associated with standard

English, though in particular countries there are some accents often

regarded as more appropriate to standard English than others. And

STANDARD LANGUAGE

272

standard English itself is not quite uniform: for example, there are

detectable differences in vocabulary and grammar between standard

American and standard British English, though these are not large.

Further, of course, standard English is not immutable: it changesfrom generation to generation, and the standard English of the

eighteenth century already sounds quaint and distant to us. Finally,

standard spoken English is not always identical to standard written

English: we write Many students are ill, but we don’t say this;

instead, we say Lots of students are ill.

No such thing as standard English existed 500 years ago; instead,

there were only innumerable regional and social varieties. But, since

then, a range of social and political pressures have combined togive rise to a single variety of English accepted as the standard

everywhere.

Standard English arose largely out of a series of countless his-

torical accidents: certain words, forms and usages happened to be

accepted by educated people in positions of prestige, while others

were not accepted. The single most important factor was the poli-

tical, economic and cultural influence of the city of London and the

surrounding region: it was largely the forms used in this area thatbecame the basis of standard English.

In recent years, there has been a vigorous and surprising debate

among academics and educators about the place of standard Eng-

lish in education: one group sees a command of standard English

as an enormous benefit and as the right of every pupil; the other

sees standard English as the elitist possession of a privileged class

and interprets attempts at teaching standard English, together with

the associated dismissal of non-standard forms of English, as hos-tile attempts at oppressing working-class speakers and denying

them power.

Almost every other language which is used in education likewise

has a recognized standard form: French, German, Finnish, Basque,

Arabic, and so on. A few languages even have two standard forms:

Norway recognizes two standard forms of Norwegian, and the

standard Mandarin Chinese of Taiwan is somewhat different from

that of China. For many years the Dutch-speakers of Belgiumattempted to maintain a different standard form of Dutch from

that used in the Netherlands, but they have now abandoned this

policy.

See also: Black English; dialect; national language; official language;vernacular

STANDARD LANGUAGE

273

Further reading: Bex and Watts 1999; Blake 1996; Crystal 1995; Holmes1992; Honey 1997; Leith 1997; Milroy and Milroy 1993, 1997.

STEM

A linguistic form which cannot stand alone but which serves as a

basis for constructing word-forms which can stand alone. In a

number of European languages, the grammatical forms of a single

word are constructed in a very orderly way; this is particularly

noticeable with verbs.

Consider Spanish. The simplest possible form of a Spanish verbis its root, a form which cannot stand alone but which is always

present in every form of that verb. In order to construct a verb-

form, we must first add one of several possible suffixes to the root;

the result, which still cannot stand alone, is one of the several stemsof that verb. Finally, to the stem is added one of several possible

endings, to produce a complete verb-form.

The root of the verb meaning ‘sing’ is cant-. To this we can add

any of several tense or mood suffixes. The present-tense suffix hap-pens to be -a-, and so the present stem is canta-. To this we can add

any of several endings to get forms like canta ‘s/he sings’ with a

zero-suffix, cantamos ‘we sing’ and cantan ‘they sing’. For the

imperfect, the suffix is -aba-, and so the imperfect stem is cantaba-,

leading to such forms as cantaba ‘s/he was singing’, cantabamos ‘we

were singing’ and cantaban ‘they were singing’. The future suffix is

-ar-, and so the future stem is cantar-, producing cantara ‘s/he will

sing’, cantaremos ‘we will sing’ and cantaran ‘they will sing’. Like-wise, the conditional suffix -arı- leads to a stem cantarı- and forms

like cantarıa ‘s/he would sing’, cantarıamos ‘we would sing’ and

cantarıan ‘they would sing’, and present subjunctive -e- yields forms

like cante ‘s/he might sing’, cantemos ‘we might sing’ and canten

‘they might sing’.

Not all languages exhibit this kind of structure (English doesn’t),

but, for those that do, recognizing roots and stems provides an

elegant account of much of their morphology.

See also: morphology; root

Further reading: Bauer 1988; Katamba and Stonham 2006; Matthews1991.

STEM

274

STEREOTYPE

In sociolinguistics, a stereotype is a linguistic feature which speakers

are so aware of as a distinctive feature that it has taken on an iconic

status for them. In other words, it is a marker which is highly sali-

ent for the speech community as a sign of identity. Stereotypesoften have the appearance of caricatures and are in fact often his-

torically outdated features which a speech community has not in

fact used for a while.

In grammar, a stereotype is a sequence of words which appears to

be made up of separate lexemes but in fact is treated as a formulaicsequence. As such, it is unproductive of further variation. Proverbs,

quotations, aphorisms and other idioms might be seen as stereo-

types in this sense.

See also: indicator; marker

Further reading: Chambers 2003; Llamas et al. 2007.

STIMULUS-FREEDOM

Our ability to say anything at all, including nothing, in any situa-

tion. The signals used by non-human species are overwhelmingly

stimulus-bound: each signal is produced always and only when a

particular stimulus is received. For example, a vervet monkey

always gives an eagle warning call upon spotting an eagle, and itnever does this at any other time.

Human language is utterly different. Save only for our (non-

linguistic) cries of pain and fear, we are perfectly free to choose

what we are going to say in every context. In other words, our

speech is stimulus-free. If your friend Julia asks What do you think

of my new skirt?, you may reply It’s too short, or It doesn’t go

with your pink blouse, or even If that doesn’t get Mike’s attention,

nothing will. You can even decline to answer, and change thesubject. Of course, there are all sorts of social pressures that

make some replies more likely than others. If you value Julia’s

friendship, you probably won’t reply God, Julia – with your fat

legs, you should stick to kaftans. But there’s nothing about English

that prevents you from saying this, and you could say it, if you

wanted to.

STIMULUS-FREEDOM

275

Though he was not the first to notice it, the American linguist

Charles Hockett drew particular attention to stimulus-freedom in

1960.

See also: design features; displacement; open-endedness


STRESS

Strong prominence on a particular syllable. In English, almost everyword has at least one syllable which is noticeably more prominent

than the other syllables; this is the stressed syllable, and the position

of the stress is usually easy to identify. All of us agree without hes-

itation that the stress falls upon the first syllable of victim and ter-

rify, on the second syllable of linguistics and invention, on the third

syllable of kangaroo and circulation, and on the only syllable of cat

and smile. There is some variation among speakers: some people

stress the first syllable of exquisite, controversy, harass and vagaries,while others stress the second.

A word of several syllables often has, in addition to its mainstress (or primary stress), a secondary stress (a lesser degree of

stress) on another syllable. For example, education has its main

stress on the third syllable but a secondary stress on the first sylla-

ble. Even short words may bear secondary stress: the words base-

ball, borax, textbook, croquet and ice cream all bear main stress on

one syllable and secondary stress on the other, though with the lasttwo examples speakers vary as to which is which.

Stress differences may distinguish pairs of words otherwise iden-

tical, as with the nouns ‘subject and ‘record and the verbs sub‘ject

and re‘cord. For some (not all) speakers, the same stress difference

is all that distinguishes billow from below.

Stress on a syllable in English is produced by a complex interac-

tion of several phonetic factors: a stressed syllable is louder than an

unstressed one, it has a higher pitch, and it is longer.Quite apart from the ordinary word-stress just described, English

also allows sentence-stress: strong stress placed on a particular word

to emphasize it within an utterance, as in I will NEVER give in to

these threats. A sentence-stress is always placed on a syllable which

bears the main stress normally. But very different is contrastivestress, in which any syllable at all, even one which is normally

STRESS

276

unstressed, may be stressed in order to highlight a contrast with

something else: I said ACcept, not EXcept.

Many other languages also have word-stress, such as Spanish,

German and Russian (and are thus termed stress-timed languages).Others (syllable-timed), however, do not: French and Japanese, for

example, lack word-stress. Some varieties of English, like Indian

English and Jamaican English, are relatively syllable-timed rather

than stress-timed like British English or Australian English.

See also: suprasegmental

Further reading: Ball and Rahilly 1999; Collins and Mees 2003; Gie-gerich 1992; Gussmann 2002; Knowles 1987; Kreidler 1989; Laver1994; Lehiste 1970; Roach 1991.

STRUCTURALISM

An approach to the study of language which sees a language as a

structured system. Before the twentieth century, linguists took anatomistic view of language: they saw a language as essentially a

collection of individual elements, such as speech sounds, words and

grammatical endings. At the beginning of the twentieth century, the

Swiss linguist Ferdinand de Saussure put forward a very different

view: he argued that a language is best viewed as a structured

system, with each element in it defined chiefly by how it is related

to other elements. In this view, which has come to be called struc-turalism, it is the system which is the primary object of study, andnot the individual elements present in that system.

Saussure’s influence helped to make structuralism the dominant

approach in European linguistics. In the USA, structuralist ideas

were somewhat independently developed by Edward Sapir and

more especially by Leonard Bloomfield. Bloomfield’s successors in

the 1940s and 1950s took his ideas to extremes in developing

American structuralism, a vigorous but excessively dogmatic

approach to linguistic description which attached great importanceto distribution. In the 1960s, Noam Chomsky and his followers

rebelled against the excesses of their predecessors, and they came to

use structuralism as a swear word to denote everything they disliked

about the earlier work, but in fact Chomsky’s generative grammar is

no less structuralist than other approaches, in the original sense of

the term. Indeed, virtually all serious work in linguistics in the

STRUCTURALISM

277

twentieth century has been structuralist in outlook, though many

contemporary linguistists continue to regard structuralism as a

term of abuse and would not apply the term to their own work.

Structuralist ideas were eventually picked up, largely from theinfluential Russian linguist Roman Jakobson, by the French

anthropologist Claude Levi-Strauss, who introduced them into

anthropology, from where they spread into the social sciences gen-

erally, and even into such fields as literary criticism. Quite inde-

pendently, Saussure’s structuralist approach was deeply influential

in the development of semiotics.

See also: cognitive linguistics; generative grammar; linguistics

Further reading: Lepschy 1970; Malmkjær 2004; Sampson 1980; Stur-rock 1993.

STRUCTURE

A particular pattern which is available in a language for construct-ing a linguistic unit, or an instance of this. Structures can be

recognized in languages at every level of analysis: phonemes com-

bine to build morphemes, morphemes combine to build words,

words combine to build phrases, phrases combine to build clauses

and sentences, sentences combine to build texts, and so on. At

every one of these levels, the smaller units must be combined into

larger ones in particular orderly ways determined by the rules of the

language, and we therefore say in each case that we are looking atan instance of a particular structure.

For example, the morpheme bad is built up from the three pho-

nemes /b/, /æ/ and /d/, and many analysts would argue that, in fact,

this is done by first combining /æ/ and /d/ into /æd/, and then

adding /b/ to produce /bæd/. The adjective happy can take the prefix

un- to produce the adjective unhappy, and this in turn can take the

suffix -ness to produce the noun unhappiness. (We cannot analyse

unhappiness as consisting instead of un- plus happiness, becausehappiness is a noun, and the rules of English word-structure do not

permit un- to be added to a noun.) The words little and girl can be

combined to construct the N-bar little girl, which can then take a

determiner like the to build the noun phrase the little girl.

Most usually today, we apply the term structure both to a general

pattern and to any individual instance of it, but the general pattern

STRUCTURE

278

is sometimes called a construction, while an individual instance has

sometimes been called a syntagm. The relation between the ele-

ments in a structure is a syntagmatic relation.

The approach to language study called structuralism gets itsname because it emphasizes the importance of recognizing units of

structure at every level, though in fact the recognition of systems in

languages is no less important in the structuralist approach.

See also: structuralism; syntagmatic relation; system


STRUCTURE-DEPENDENCE

The property of languages by which grammatical statements must

be made in terms of structural units. When we describe the syntax

of a language, we find that we never need to make any statements

involving such notions as ‘the second word in the sentence’, ‘a word

ending in a vowel’, ‘a word of three or more syllables’ or ‘the firstsix words in the sentence’. Grammars simply do not function in

terms of such notions; instead, they function in terms of purely

structural units like verb, noun phrase and auxiliary, and we call this

phenomenon structure-dependence.Consider the way in which yes–no questions are constructed in

English. If we mentally start with a statement, then, in order to

convert it to a yes–no question, we must move the finite auxiliary

out of its verb phrase and to the left of the subject noun phrase.This is so regardless of what the finite auxiliary is and regardless of

whether the noun phrase and the verb phrase are small and simple

or large and complex:

[She] [[is] [dancing]].

[Is] [she] [dancing]?

[Most of our students] [[can] [do this]].[Can] [most of our students] [do this]?

[The selection of the remaining members of the team] [[can’t]

[wait until our captain gets back]].

[Can’t] [the selection of the remaining members of the team]

[wait until our captain gets back]?

STRUCTURE-DEPENDENCE

279

In short, yes–no questions are constructed in terms of structural

units, not in terms of words, and every other syntactic process in

English and in other languages works in the same way,

It is because of this structure-dependence that descriptions of thesyntax of a language are commonly presented in terms of con-

stituent structure and in terms of syntactic categories.

See also: constituent structure; recursion; structure; syntactic category

Further reading: Jeffries 2006; Radford 1988.

STYLISTICS

The study of the aesthetic uses of languages, particularly the use of

language in literature. On the whole, the European linguistic tradi-

tion has almost always seen the study of the purely structural

aspects of language as bound up with the study of its aesthetic

aspects; for example, in the middle of the twentieth century, the

great Russian linguist Roman Jakobson contributed equally totheoretical linguistics and to the critical examination of literary

works.

In the English-speaking world, however, there was long a see-

mingly unbridgeable gulf between linguistics and literary criticism,

and neither discipline paid any attention to the other. In the past

several decades, however, this has changed, and a number of scho-

lars have been applying the analytical techniques of theoretical lin-

guistics to the elucidation of literary works and to the examinationof the aesthetic aspects of language generally. To this discipline we

now give the name stylistics. A practitioner of stylistics may choose

to pursue such topics as the ironic use of language in the novels of

V.S. Naipaul, the sound structure of a poem by Dylan Thomas, the

use of regional and social varieties of English for comic effect in

Shakespeare’s Henry V, or the use of archaisms to achieve stateli-

ness in religious language. What sets all this apart from ordinary

literary criticism is the explicit use of the concepts and analyticaltechniques of linguistic theory. As linguistics (and psychology,

social theory and the philosophy of language) has developed, so

stylistics has borrowed its insights in order to augment its analytical

toolkit. For example, early stylistics in the 1960s and 1970s focused

on semantic and syntactic aspects of poetry, largely; the 1980s saw

an expansion drawing on pragmatics and discourse analysis into

STYLISTICS

280

extended prose fiction and drama; the 1990s and recent work has

drawn heavily on cognitive science to produce a cognitive poeticsand on computational linguistics to produce a corpus stylistics.

In the past few years, the label literary linguistics (once used inthe 1920s in Russia) has begun to be applied again to the linguistic

analysis of literature.

See also: cognitive linguistics; structuralism

Further reading: Bradford 1997; Fabb 1997; Fowler 1996; Simpson1996, 2004; Stockwell 2002b; Thornborrow and Wareing 1998; Wales2001.

SUBCATEGORIZATION

Differences in syntactic behaviour among the words in a single part

of speech. English has a class of verbs, almost all of which are

united by sharing certain important properties, such as being

marked for tense and taking the ending -ing for certain grammaticalpurposes. If we didn’t recognize a class of verbs, we would be

unable to state a number of important generalizations. Never-

theless, not all verbs behave identically.

Consider the following frames:

1 Susie will –.

2 Susie will – Mike.

3 Susie will – me a letter.4 Susie will – to come.

5 Susie will – that she’s ready.

6 Susie will – me that she’s ready.

Now try to fit the following verbs into each blank: decide, tell, want,

send, speak, show. You will find that each verb fits successfully into

some blanks to produce a grammatical sentence, but not into

others. For example, decide fits only into (1), (4) and (5).This is an instance of subcategorization. Even though all these

words are verbs, they exhibit important differences in the structures

in which they can appear, and we say that verbs are subcategorizedaccordingly. Note that the necessary subcategories all overlap con-

siderably, and hence we cannot simply subcategorize verbs into type

(1), type (2), and so on.

SUBCATEGORIZATION

281

Subcategorization is most prominent with verbs, but other parts

of speech show the same phenomenon.

See also: part of speech; syntactic category; syntax

Further reading: Radford 1988.

SUBJECT

The subject is the primary argument of a sentence. Traditionally, the

subject is the nominated actor or ‘doer’ of the verb, though manygrammars distinguish the grammatical subject from the logical or

‘real’ subject: in The baby was kissed by the politician, the gramma-

tical subject is different from the logical subject. And there are

examples where the subject is plainly not the active agent: the lake

shimmered in the wind, The wine drinks well. The identification of

subject is thus relatively straightforward at the surface level, but

more difficult when function is considered. It should be said,

though, that all languages have subjects, though they encode thenomination of the primary argument in various different ways.

A subject is prototypically a noun phrase, though it is possible for

other parts of speech and categories to act as subjects: verb phrases

can be nominalized or rank-shifted to operate as subjects (Running

up the hill is tiring, To finish this book is his main aim). Some lan-

guages allow a dummy subject or expletive (in the grammatical

sense) where the syntax requires a subject-slot to be filled but the

pronoun used has no semantic referential value. Typical in Englishare existential statements (There are ten boys here), statements that

introduce modal forms (It is true that there are no bananas) or

assertions about the weather (It is raining).

See also: argument; direct object; noun; noun phrase


SUBORDINATION

Any type of sentence structure in which one clause forms part of a

larger clause. The recognition of subordination is ancient in our

grammatical tradition, and any clause which is part of a larger one

SUBJECT

282

is a subordinate clause. Subordinate clauses are of several types; in

the examples below, each subordinate clause is bracketed.

A complement clause is attached to a preceding word: Janet sus-

pects [that she is pregnant] (a verb-complement clause); Susie’s

announcement [that she was a lesbian] startled her parents (a noun-complement clause); Susie is happy [that Natalie can move in with

her] (an adjective-complement clause).A relative clause is attached to a preceding noun, which it modi-

fies, and always contains a gap (e) somewhere inside it: The job [that

Susie wants e] involves a lot of travelling.

An embedded question (or indirect question) is a question which is

not being asked directly: I asked them [if they had seen Susie]; Janetdoesn’t know [what she wants].

An adverbial clause behaves like an adverb within its sentence and

must be introduced by a suitable marker of subordination: [When

Susie gets here], we’ll have dinner; [If you need any help], give me a

ring; Susie is upset [because she has quarrelled with Natalie].

Earlier grammarians took the view that a subordinate clause was

not part of the clause containing it (its matrix clause), but today we

always regard a subordinate clause as forming an integral part ofits matrix clause.

See also: clause; sentence

Further reading: Collins Cobuild 1990; Greenbaum and Quirk 1990;Hurford 1994.

SUPRASEGMENTAL

An aspect of pronunciation whose description requires a longer

sequence than a single consonant or vowel. Though phoneticians

and linguists had earlier been aware of the importance of supra-

segmental phenomena in speech, the term suprasegmental was

coined by the American structuralists in the 1940s. It covers several

rather diverse phenomena.A very obvious suprasegmental is intonation, since an intonation

pattern by definition extends over a whole utterance or a sizable

piece of an utterance. Also clearly suprasegmental is the kind of

word-accent called pitch accent, as in Japanese and Basque, in

which an accentual pitch pattern extends over an entire word. Less

obvious is stress, but not only is stress a property of a whole syllable

SUPRASEGMENTAL

283

but the stress level of a syllable can only be determined by com-

paring it with neighbouring syllables which have greater or lesser

degrees of stress.

The tones of tone languages are also suprasegmental, since notonly does a tone fall on a whole syllable but tonal differences like

‘high’ and ‘low’ can only be identified by comparing syllables with

neighbouring syllables.

The American structuralists also treated juncture phenomena as

suprasegmental. Differences in juncture are the reason that night

rate does not sound like nitrate, or why choose like white shoes, and

why the consonants in the middle of pen-knife and lamp-post are

the way they are. Since these items contain essentially the samesequences of segments, the junctural differences have to be descri-

bed in terms of different juncture placement within sequences of

segments.

In most of these cases, the phonetic realization of the supraseg-

mental actually extends over more than one segment, but the key

point is that, in all of them, the description of the suprasegmental

must involve reference to more than one segment.

Some people use prosody as a synonym for suprasegmental, butphonologists prefer the latter term.

See also: intonation; segment; stress; tone language

Further reading: Ball and Rahilly 1999; Clark and Yallop 1995; Lade-foged 1993; Laver 1994; Lehiste 1970; Shockey 2003.

SURFACE STRUCTURE

The syntactic structure which is most obviously assignable to a

particular sentence. Every sentence in every language has some kind

of syntactic structure. Consider the following English example:

Susie is hard to please. Now, while the structure assigned to this

sentence may vary somewhat, depending upon the analyst and the

framework used, the majority view would probably analyse it asfollows: it consists of a subject noun phrase Susie and a predicate

verb phrase is hard to please; the latter in turn consists of a copular

verb is and an adjective phrase hard to please; the adjective phrase in

turn consists of an adjective hard and a verb phrase to please. We can

represent this structure schematically as follows: [Susie] [is] [hard]

[to please], or even better, given more space, as a tree diagram.

SURFACE STRUCTURE

284

Now this structure is the surface structure of the sentence: that is,it is the syntactic structure which we would most naturally assign to

this sentence, and, in many theories of grammar, this is the only

syntactic structure that would be assigned. However, it is obviousthat the meaning of the sentence relates to this structure in a very

curious way: this is really a statement about pleasing Susie, and

hence, in some semantic or logical sense, Susie appears to be, not

the subject of is, but rather the object of please.

Most frameworks would treat this fact in a strictly semantic

way, but transformational grammar and its various descendants

take a different line: they assign a deep structure to the sentence,

a structure which is very different from the surface structure andmuch closer to the meaning – something along the following

lines: [[NP] [please] [Susie] [is] [hard], where ‘NP’ represents an

unidentified subject for please. A good deal of grammatical

machinery is then required to convert this deep structure into the

required surface structure. These rules are transformations, and it

is the set of transformations that define languages as being dif-

ferent from each other, producing different apparent surface

structures.

See also: deep structure; tree

Further reading: Lyons 1991.

SYLLABLE

A fundamental but elusive unit in phonology. Every word consists

of a sequence of some number of syllables, and even speakers with

no knowledge of phonetics or of linguistics usually find it easy to

agree on how many syllables a word contains. For example, we all

agree that girl and salt contain one syllable, that butter and behind

contain two, that linguistics and kangaroo contain three, that edu-

cation and development contain four, and so on.

Differences in judgement usually reflect genuine differences inpronunciation. For example, police and collapse are pronounced as

one syllable by some speakers but as two by others, library and

medicine are pronounced as two syllables by some but as three by

others, February is often uttered with two syllables and can be read

out with four, and temporarily has five syllables for most Americans

but only three for many Britons.

SYLLABLE

285

In spite of the conspicuous prominence of syllables, however,

it has proved exceedingly difficult to provide an objective defi-

nition of the syllable. Attempts have been made at defining the

syllable in terms of everything from muscular contractions toperceived peaks of loudness, but so far no definition has proved

satisfactory.

Nevertheless, the syllable appears to be a genuinely fundamental

unit in phonology, and many important generalizations about the

sound systems of languages can be stated in terms of syllables more

readily than in any other way. As a consequence, the most influen-

tial theories of phonology developed since the 1980s have generally

attached great importance to syllables, and in some of them thesyllable is taken as the most fundamental unit of all, in terms of

which everything else is defined. (In contrast, earlier theories of

phonology generally took either phonemes or distinctive features as

the fundamental units).

See also: phonology; segment

Further reading: Gussmann 2002; Hogg and McCully 1987; Katamba1989; Lass 1984; Odden 2005.

SYMBOLIC SYSTEM

An integrated set of signs, each with a conventional meaning. The

notion of a symbolic system lies at the heart of the discipline called

semiotics. A very simple symbolic system is the set of coloursexhibited by a traffic light, each of which has a conventional but

agreed significance. Only slightly more complex is the set of signals

used by a cricket umpire or an American football referee to

announce what has happened on the field. Yet another example is

the system of stripes, bars, stars and other items worn on a military

uniform to indicate rank and affiliation.

Language is often viewed as a paradigm case of a symbolic

system, though one that is vastly more complex than these simpleexamples. Much of the driving force of semiotics is the belief that

such socially constructed objects as myths and films can also be

usefully viewed as symbolic systems, and interpreted as such in

illuminating ways.

See also: linguistic sign; semiotics

SYMBOLIC SYSTEM

286

Further reading: Cobley 2001; Eco 1976; Leeds-Hurwitz 1993; Peirce1992; Sebeok 1984, 1994; Tobin 1990.

SYNCHRONY

The absence of a time element in linguistic description. In the early

twentieth century, the Swiss linguist Ferdinand de Saussure intro-

duced his celebrated distinction between synchrony and diachrony.

In a synchronic approach to describing a language, we focus on

that language at one moment in time and describe it as we find it at

that moment (a sort of ‘snapshot’ of the state of the language). Thisneed not be the present moment: we can equally construct a

description of present-day English or of Shakespeare’s English. In

either case, we take no interest in how the language of that moment

differs from the same language at any earlier or later moment; as

soon as we start paying attention to that, we are taking a dia-

chronic approach. In linguistic research, synchronic studies are

obviously much more directly evidential than diachronic studies,

which tend to require complex methodologies and inferencing.

See also: descriptivism; diachrony

Further reading: Nevalainen et al. 2006.

SYNTACTIC CATEGORY

Any one of the several types of grammatical unit from which the

sentences of a language are constructed. The structure of a sentence

is constituent structure, and a constituent structure is built up by

combining smaller units into larger ones. These units, the syntacticcategories, are of several types, and they are also of different sizes.

The smallest syntactic categories are the lexical categories, com-

monly called the parts of speech, such as noun, verb, preposition and

determiner; there are about fifteen of these in English. Eachmember of a lexical category is a single word.

Much larger are phrasal categories, representing the several dif-

ferent types of phrase existing in the languages, such as noun phrase

and verb phrase. A typical instance of a phrasal category consists of

several words, but a particular phrase may contain only one word

or dozens. Most grammarians recognize only five different phrasal

SYNTACTIC CATEGORY

287

categories in English, though some prefer to acknowledge a larger

number.

In between these two sizes are the intermediate categories. Theneed for these is debated among grammarians, but one kind isessential: the N-bar. A noun phrase like the little girl unquestion-

ably consists of a determiner the and an N-bar little girl, which

itself consists of the adjective little and the noun girl.

Of course, we also need the unique syntactic category sentence.

Subordinate clauses present a few difficulties. In the sentence Susie

decided that she would go home, the complement clause that she

would go home is often assigned to another special syntactic cate-

gory called S-bar; an S-bar consists of a complementizer (here that;complementizer is another part of speech) plus the sentence she

would go home.

The recognition of a suitable set of syntactic categories allows us

to analyse all the sentences of a language as being built up, by

means of a fairly small set of rules allowing recursion, from just

these few categories.

See also: head; part of speech; phrase; recursion; subcategorization;X-bar

Further reading: Brown and Miller 1991; Kroeger 2004; van Valin 2001.

SYNTAGMATIC RELATION

The relation between any linguistic elements which are simulta-neously present in a structure. The concept of a syntagmatic relationwas introduced in the early twentieth century by the Swiss linguist

Ferdinand de Saussure. In Saussure’s terminology, any kind of

structural unit in a language at any level is a syntagm or syntagma,a word constructed on the analogy of paradigm, and a syntagm

consists of some set of smaller structural units, all combined

according to the appropriate rules and all standing in syntagmatic

relation to one another.For example, the English phonemes /t/, /k/ and /æ/ are syntag-

matically related in one way in the word /tæk/ tack, in a second way

in /kæt/ cat, in a third way in /ækt/ act, and in a fourth way in

/tækt/ tact. Similarly, the morphemes {un-}, {happy} and {-ness}

are syntagmatically related in a particular way in the word unhap-

piness, and the words the, little and girl are syntagmatically related

SYNTAGMATIC RELATION

288

in the noun phrase the little girl. In every case, the larger unit which

is composed of smaller ones is a syntagm.

A particular pattern into which a large number of individual

syntagms fit is a structure. So, for example, the structure illustratedby the little girl recurs in other, similar, syntagms like these old

clothes, my new car and some dirty books. (Today, however, we more

usually apply the term structure to each particular instance of a

general pattern, and avoid the word syntagm).

Syntagmatic relations contrast most obviously with paradigmatic

relations, and this pair of terms constitutes one of Saussure’s

famous dichotomies.

Note that, in psycholinguistics, the term syntagmatic relation isused in a somewhat different sense, to denote the mental associa-

tion between words which frequently occur together, as when a

subject given black responds with magic, tie or sheep.

See also: distribution; paradigmatic relation; structure

Further reading: Kroeger 2004; van Valin 2001.

SYNTAX

Sentence structure, or the branch of linguistics which studies this.

The first European steps in the examination of syntax were taken

by the ancient Greeks, beginning with Aristotle, who first divided

sentences into subjects and predicates. Thereafter, progress was slow,

and toward the middle of the twentieth century syntax was laggingfar behind phonology and morphology.

In the 1940s, the American linguist Zellig Harris began develop-

ing an interesting new way of looking at syntax. A decade later, his

student Noam Chomsky presented a greatly modified version of

Harris’s ideas, involving the introduction of generative grammar and

of the particular variety of generative grammar called transforma-

tional grammar. Chomsky argued that syntax was not only tractable

but the very heart of serious linguistic investigation, and he per-suaded a generation of linguists that he was correct. As a result, the

study of syntax became vastly more prominent than formerly; still

today, many linguists of a Chomskyan persuasion see syntax as the

very core of language structure, though non-Chomskyan linguists,

thanks to the dramatic advances in other areas of investigation,

would now see syntax as only one important area among many.

SYNTAX

289

In the 1960s, the American linguist Joseph Greenberg published

his pioneering work in syntactic typology, as a result of which another

tradition of syntactic investigation, independent of Chomsky’s, has

grown up and flourished. Often closely linked to this typologicalwork, but partly independent of it, is the kind of syntactic work

favoured by the proponents of functionalism.

This recent concentration of effort upon syntactic problems has

uncovered a wealth of fascinating data and led to innumerable the-

oretical interpretations. Here are just two examples of the many

striking phenomena discovered in English.

First, consider the following four virtually identical-looking

sentences:

1 After Lisa got up, she had a shower.

2 After she got up, Lisa had a shower.

3 Lisa had a shower after she got up.

4 She had a shower after Lisa got up.

In the first three of these, she can possibly refer to Lisa, but, in the

fourth, it cannot.Second, the sentence It is easy to annoy Janet can be readily

recast as Janet is easy to annoy, but the similar-looking It is inad-

visable to annoy Janet cannot be recast as *Janet is inadvisable to

annoy (*indicates ungrammaticality).

Facts like these, previously unnoticed and largely unsuspected,

have provided the grist for decades of syntactic investigation and

theorizing, and quite a number of different theories of grammar havebeen put forward and developed.

See also: recursion; structure-dependence; syntactic category

Further reading: Brown and Miller 1991; Givon 2001; Hendrick 2003;Hudson 1998; Jeffries 2006; Kroeger 2004, 2005; Kuhn 2003; Lasnik etal. 2005; Lyons 1968; Matthews 1981; O’Grady et al. 1996; Poole 2002;van Valin 2001.

SYSTEM

A set of competing possibilities in a language, together with the

rules for choosing among them. The single greatest insight of the

approach to language study called structuralism was the recognition

SYSTEM

290

that a language is best viewed as a system of elements, with each

element being chiefly defined by its place within the system, by the

way it is related to other elements. Further, the overall system of a

language consists of a number of subsystems and sub-subsystems,all of which overlap in various ways.

A simple example of a system in English is our (personal) pro-

noun system, which consists of the items I/me, we/us, you, he/him,

she/her, it and they/them. Whenever we need to use a pronoun, we

must choose exactly one item from this system, and in many cases

we must choose one of the two forms that exist. Naturally, the

choice is not free: it is dependent upon other choices which are

being made at the same time. This whole set of forms, together withthe rules determining which one is appropriate in a given instance,

constitutes our pronoun system.

Among the other systems present in English are the consonant

system, the verb system, and the system of word-forming prefixes.

All the items forming part of a single system stand in a paradig-

matic relation to one another, meaning that they represent mutually

exclusive choices.

See also: paradigmatic relation; structuralism; structure

Further reading: Thompson 2004.

SYSTEMATIC CORRESPONDENCE

A certain type of pattern linking words in different languages. Themembers of a language family share a common ancestor – that is, they

all started off long ago as nothing more than regional dialects of that

ancestral language, but have diverged over time into distinct languages.

One type of change is change in pronunciation, and it is characteristic

of change in pronunciation that it is often highly regular: a parti-

cular sound in a particular environment tends strongly to change in

the same way in a given language in every word containing it.

As a result of this, the words of related languages often exhibit a setof conspicuous patterns, each of the following general form: if word

W1 in language L1 contains a sound S1 in a particular position, then

word W2 of the same meaning in language L2 will contain the sound

S2 in the same position. Such a pattern is a systematic correspondence.Here is a simple example, from the two New Guinean languages

Sinaugoro and Motu (data from Terry Crowley 1997); these data

SYSTEMATIC CORRESPONDENCE

291

are only a sample of some patterns which in fact apply to a much

larger set of words:

We can easily see several systematic correspondences here which

apply in all positions: Sinaugoro /m/ : Motu /m/; Sinaugoro /n/:

Motu /n/; Sinaugoro /a/ : Motu /a/; Sinaugoro /¥/ : Motu zero; and

so on. Whenever a Sinaugoro word contains one of the sounds in a

correspondence, the Motu word of the same meaning contains the

other one in the same position.

But Sinaugoro /t/ is more complicated: some words show thecorrespondence Sinaugoro /t/ : Motu /t/, while others show Sinau-

goro /t/ : Motu /s/. Further examination, though, reveals that there

is a clear basis for the difference: the second correspondence applies

always and only when the consonant is followed by a front vowel /i/

or /e/, while the first applies in all other circumstances.

The existence of such correspondences demonstrates that the

languages in question share a common ancestor (in this case, of

course, a rather recent one), and these correspondences are the mate-rial to which we apply comparative reconstruction in order to work

out the properties of the unrecorded ancestral language.

See also: comparative reconstruction; historical linguistics; language family

Further reading: Crowley 1997; Trask 1996.

SYSTEMIC LINGUISTICS

An important version of functionalism. In the 1930s and 1940s, theBritish linguist J.R. Firth began laying the groundwork for a

Sinaugoro Motu

Tama tama ‘father’Tina sina ‘mother’Tai tai ‘cry’Tui tui ‘elbow, knee’¥ita ita ‘see’¥ate ase ‘liver’Mate mase ‘die’Natu natu ‘child’Toi toi ‘three’


292

somewhat novel social approach to language. His student Michael

Halliday greatly developed Firth’s ideas in distinctive directions

of his own. Beginning in the 1960s with a new approach to

grammatical analysis which he called Scale-and-Category Gram-mar, Halliday went on to construct an elaborate and ambitious

framework which eventually came to be called Systemic Linguistics,or SL.

SL is an avowedly functionalist approach to language, and it is

arguably the functionalist approach which has been most highly

developed. In contrast to most other approaches, SL explicitly

attempts to combine purely structural information with overtly

social factors in a single integrated description. Like other func-tionalist frameworks, SL is deeply concerned with the purposes of

language use. Systemicists constantly ask the following questions:

What is this writer (or speaker) trying to do? What linguistic devi-

ces are available to help them to do it, and on what basis do they

make their choices?

Halliday distinguishes among three distinctive functions of lan-

guage (or metafunctions). The ideational (or experiential) function is

the conveying of semantic content representing information aboutour experience of the external world (including our own minds).

The textual function is the linking of linguistic elements to other

linguistic elements, so that the various parts of a text can be inte-

grated into a coherent and cohesive whole and related to the wider

context of our speech or writing. The interpersonal function is the

establishment and maintenance of social relations, including per-

suading other people to do things or to believe things.

Systemicists stress the utility of their framework in the analysis oftexts, an area beyond the scope of many other approaches, and they

accordingly devote more attention to the treatment of texts than to

the analysis of isolated sentences. Because of this preoccupation

with texts, the concepts of coherence and cohesion play a central

role in the framework. And SL has proven useful especially in the

fields of stylistics and critical discourse analysis.

Halliday and his followers have recently been applying the name

Functional Grammar to the more explicitly grammatical aspects ofSL, and the term systemic functional linguistics has also been used.

See also: cohesion; functionalism; text

Further reading: Bloor and Bloor 2004; Butler 1985; Eggins 2004;Halliday 2004; Thompson 2004.


293

TAGMEMICS

A grammar developed by the American linguist Kenneth Pike in

the 1950s. In tagmemic theory there are three main levels or sys-tems: phonology, lexis and grammar and within each level there are

sub-levels. So in tagmemic grammar there are morpheme, word,

phrase, sentence and paragraph levels. Tagmemics maps the same

distinction between the abstract phoneme and the concrete pho-

netic sound onto higher levels of language. So a set of synonyms

(letter, note, line, epistle, message, missive, correspondence) can be

regarded as different realizations of the single conceptual tagmemeunderlying them. At a syntactic level, for example, subject, objectand predicate tagmemes are realized by surface words. Pike was the

first to make this emic/etic distinction (derived from phonemic/pho-

netic), where an emic analysis refers to the significant concept that

the researcher is interested in, and the etic aspect is the actual rea-

lization that would be familiar to the language users.

See also: grammar

Further reading: Pike 1967, 1982.

TENSE

The grammatical category which relates to time. Every language is

capable of expressing limitless distinctions of time: soon, tomorrow,

next Wednesday at 2.00, 137 years ago, 138 years from now. It ispossible for a language to build a few of these time distinctions into

its grammar, and a language which does so has the category of

tense. Tense is thus the grammaticalization of time. In most tense

languages, tense is marked on verbs, but there are exceptions.

Some languages lack tense entirely; an example is Chinese, which

has nothing corresponding to the I go/I went contrast of English.

Some tense languages distinguish only two tenses, while others have

three, four, five or more; the African language Bamileke-Dschangdistinguishes eleven tenses. Other possibilities for tenses include

distinguishing near-future tense (today, or soon after today) from

distant future events, hypothetical or desired events, and events

which are still in the process of happening, as well as several other

possible time signatures. A useful way of describing tense and

aspect in discourse is to compare the relative positions of the coding

TAGMEMICS

294

time (the moment of utterance or composition) with the referencetime (the explicit moment being referred to) and the event time (the

event’s duration or status). Combining all possibilities of sequencing

of these before or after each other produces a scheme of twenty-twopossible forms, without constraints of particular languages.

English has only two tenses: a non-past (‘present’) tense, mostly

used for talking about present and future time, and a past tense,

mostly used for talking about past time. English verb-forms there-

fore generally come in pairs, one for each tense:

She lives in London/She lived in London.

She is living in London/She was living in London.She has lived in London/She had lived in London.

She has been living in London/She had been living in London.

She is going to live in London/She was going to live in London.

She will live in London/She would live in London.

She can live in London/She could live in London.

Unlike some other languages, English has no distinct future tense.

Instead, we use a variety of non-past (‘present’) forms for expressinga range of attitudes towards future events: I go to London tomorrow;

I’m going to London tomorrow; I’m going to go to London tomorrow;

I’ll go to London tomorrow; I’ll be going to London tomorrow; I must

go to London tomorrow; I may go to London tomorrow. Observe that

all these examples illustrate some further distinctions of importance

in English grammar, but these others are not distinctions of tense.

Instead, they are distinctions of aspect or of modality.

See also: aspect, modality

Further reading: Bybee et al. 1994; Comrie 1985; Hornstein 1990; Hur-ford 1994.

TEXT

A continuous piece of spoken or written language, especially one

with a recognizable beginning and ending. Linguists have long used

the word text very informally to denote any stretch of language

they happened to be interested in. Especially since the 1960s, how-

ever, the notion of a text has acquired a theoretical status in several

quarters, and the analysis of texts is now seen as a major goal of

TEXT

295

linguistic investigation. However, the conception of what con-

stitutes a text is not everywhere the same.

For some linguists, a text is no different from a discourse. For others,

a text is a more or less physical product, the result of a discourse, whichitself is then seen as a more abstract process leading to the construction

of a text. For still others, a text is primarily defined by its possession

of an identifiable purpose, an approach which leads quickly to the

classification of texts into a number of kinds (text-types) differingin purpose – and, consequently, often also in their linguistic char-

acteristics. Yet others see a text as an abstraction, with a discourse

being the physical realization of a text. Finally, some linguists

merely consider that a text is written while a discourse is spoken.The analysis of texts is a prominent feature of several types of

functionalism, and above all of Systemic Linguistics, in which the

analysis of texts is often seen as the primary goal of linguistic

investigation, with the analysis of smaller units like sentences being

interpreted largely in terms of their contribution to a text. Quite

independently, the approach to teaching English called language inuse focuses strongly upon the analysis of texts, particularly those

which are familiar and meaningful to students. In Europe, a parti-cular approach called text linguistics has become prominent in

recent decades; this lays particular emphasis on textuality, the

defining characteristics of different types of texts.

Particularly associated with Systemic Linguistics, but also pro-

minent in other approaches, are the two concepts of coherence and

cohesion. Some linguistic approaches have recently incorporated the

originally literary concept of intertextuality. Some philosophers of

language take the widest possible view that anything you canrepresent to yourself (that is, any thoughts you can consciously

have) is a text, and so in effect everything is a text.

See also: cohesion; intertextuality; Systemic Linguistics; text linguistics;textuality

Further reading: Crystal 1997; Gramley and Patzold 1992; Schiffrin1994; Thompson 2004; van Peer 1994.

TEXT LINGUISTICS

A particular approach to the analysis of texts. Text linguistics is

primarily a European creation, and it is especially prominent in

TEXT LINGUISTICS

296

Germany and the Netherlands. The approach focuses upon the

varying purposes of different texts and upon the explicit identifica-

tion of the formal linguistic properties which distinguish one type

of text from another; these properties are taken to define the tex-tuality of a text. In the 1970s, a pioneering project at the University

of Konstanz in Germany attempted to construct an explicit text

grammar; the project was not seen as a success, and more recent

investigations have been characterized by greater elaboration and

sophistication.

Text linguistics makes heavy use of familiar linguistic concepts

and terminology, and much work in the field consists of attempts at

extending familiar types of linguistic analysis to units larger than asingle sentence. Consequently, it has a great deal in common with

the approach called discourse analysis in the English-speaking

world, and some outsiders see little difference between the two. The

functionalist approach called Systemic Linguistics shares important

ideas with text linguistics, but is rather distinct in nature.

See also: discourse analysis; Systemic Linguistics; text; textuality

Further reading: de Beaugrande 1994; de Beaugrande and Dressler1981; Malmkjær 2004.

TEXT WORLD THEORY

An approach to human discourse processing developed initially by

the British linguist Paul Werth in the 1990s. Text world theoryaddresses the main problem for worlds theories which concerns how

the knowledge frame is used to select only that textual and con-

textual information which is needed for the interpretation at hand

and disregard the rest. Text world theory places text-drivenness at

the heart of the process and provides a range of semantic processes

(assertion, entailment, presupposition, inferencing, and so on) to

explain which knowledge is nominated and brought to bear upon

the discourse situation.A key useful feature of the theory is its fractal structure. The top

level of the discourse world corresponds with reality, and is built

with objects and participants. The next level, the text world, has a

different ontological status but is also built with objects, partici-

pants and a spatiotemporal signature. Further embedded levels

(sub-worlds in the early theory) are the products of world switches

TEXT WORLD THEORY

297

that are effected by, for example, modalization, metaphor, negation

or deictic shifts ‘backwards’ or ‘forwards’ – they too have the same

organizational structure. Text world theory draws on other worldstheories in its use of counterparts of elements operating acrossworlds. The theory has proven most useful in relation to literary

discourse in stylistics and cognitive poetics.

See also: cognitive linguistics; possible worlds theory; schema; stylistics

Further reading: Werth 1999; Gavins 2007; Stockwell 2002b.

TEXTUALITY

The characteristics of a text which make clear what sort of text it is

intended to be. A newspaper story does not resemble a scholarly

monograph, and a poem is quite dissimilar to a television com-

mercial. Each particular type of text has its own typical character-

istics; when we encounter a text, we expect to see the appropriate

characteristics, and recognizing those characteristics allows us torecognize quickly what sort of text we are looking at.

The identifying properties of each type of text constitute its tex-tuality, or texture (though the latter term is also used in cognitivepoetics to refer to the aesthetic experience of reading). One of the

principal goals of text linguistics is to identify, as explicitly as pos-

sible, the distinguishing features of each type of text.

See also: genre; intertextuality

Further reading: Halliday 2004; Halliday and Hasan 1976; Stockwell2007; Thompson 2004.

TONE LANGUAGE

A language in which words can be distinguished purely by the pitchof the voice used on individual syllables. In a tone language, words

consisting of identical sequences of consonants and vowels can be

distinguished in pronunciation (and meaning) by the differing ways

in which the pitch of the voice behaves on each syllable; these dif-

ferent pitch patterns are called tones. A tone language may have

between two and eight (or, rarely, more) different tones.

TEXTUALITY

298

Mandarin Chinese is a tone language with four tones. In Man-

darin, a syllable like shu means nothing until it receives a tone. We

thus have shu- ‘write’ (with a high-level tone), shu ‘sorghum’ (with a

rising tone), shu ‘technique’ (with a falling tone), and shu ‘category’(with a falling-rising tone).

Most Chinese words are one syllable long, but not all tone lan-

guages are like this. Margi (spoken in Nigeria) has contrasts like

these: yına ‘to dye’ (two high tones), yına ‘to rinse’ (two low tones),

ama ‘husband’s mother’ (high-low), ama ‘but’ (low-high).

Tones can be used for grammatical purposes. Many African lan-

guages use tones to inflect verbs, as does Kanuri (spoken in

Nigeria): leze ‘he goes’ (falling tone), leze ‘he is to go’ (high tone);tusse ‘he rests’, tusse ‘he is to rest’.

In a true tone language, every syllable has its own tone, which is

independent of the tones on other syllables. Rather different is a

language with a pitch accent, in which a single pitch contour is

superimposed upon an entire word, and the pitch of one syllable is

not independent of the pitch of other syllables. In Japanese, for

example, which has two possible pitches, high (H) and low (L), all

that can ever happen in a word is that the pitch may go up onceand then (later) go down once. Examples: hana (LL) ‘nose’, hana

(LH) ‘flower’; shiro (LH) ‘castle’, shiro (HL) ‘white’; sakura (LHH)

‘cherry’, zakuro (HLL) ‘pomegranate’, kokoro (LHL) ‘heart’. Pat-

terns like HLH, HLLH and LHLH are prohibited, and so also are

HH and HHH.

See also: intonation; suprasegmental

Further reading: Hyman 1975; Katamba 1989; Ladefoged 1971; Lehiste1970; Yip 2002.

TOPIC

That part of a sentence or utterance which the whole thing is

‘about’. The division of a sentence, from the point of view of itsinformation content, into a topic and a comment is essentially the

same as the given/new distinction, but the notion of a topic has

some further linguistic uses.

The topic of a sentence is that part of it which the whole sentence

is about. For example, if I am advising my students as to which

books they should read to learn about a particular subject, I might

TOPIC

299

hold up a particular book and say I can’t recommend this book.

Here the topic is clearly this book: this noun phrase identifies what

I’m talking about, and the rest of my utterance constitutes the

comment, what I’m saying about it.English provides us with some more explicit ways of marking

something as a topic. One is topicalization, in which the topic is

simply moved to the front of the sentence: This book I can’t

recommend. Here this book has been explicitly topicalized.

Another device is the as for construction: As for this book, I can’t

recommend it.

The notion of a topic must be clearly distinguished from that of

focus; even some professional linguists confuse these two.

See also: focus; given/new

Further reading: Brown and Miller 1991; Greenbaum and Quirk 1990;Thompson 2004.

TRADITIONAL GRAMMAR

The entire body of grammatical work done in Europe and America

before the rise of modern linguistics in the twentieth century. The

European grammatical tradition began with the Greeks and was

continued by the Romans, both of whom were chiefly interested in

describing their own languages. The descriptive procedures and

terminology they developed eventually came to be applied to modern

languages like French and English, and the resulting Latin-baseddescriptions of English were taught in what for 400 years were

called ‘grammar schools’ until at least the 1960s, since when many

schools in English-speaking countries have ceased teaching any

English grammar at all.

The work of the traditional grammarians still forms the founda-

tion of modern grammatical work, but we have introduced very

many changes and extensions. We reject their insistence on pre-

scriptivism as the basis of description; we recognize many moreparts of speech than they did; we assign constituent structure to

sentences; we have identified a large number of grammatical phe-

nomena which they overlooked; we try to construct generative

grammars; and we deny that there is anything special about the

grammar of Latin, which is now seen as just one possible gram-

matical system among many.

TRADITIONAL GRAMMAR

300

Some contemporary work, such as the series of English gram-

mars prepared by Randolph Quirk and his colleagues, is still notice-

ably traditional in orientation, but it is nonetheless strongly influenced

by the advances in grammatical study achieved during the twentiethcentury.

See also: descriptivism; linguistics; prescriptivism

Further reading: Lepschy 1994; Malmkjær 2004; Palmer 1971; Robins1997.

TRAJECTOR

A term from cognitive grammar to explain how prepositions are

conceptualized. Prepositions like over, through, into, from, against,

about and so on are metaphorically derived from an idealized spa-

tial concept which is called an image-schema. The notion of over,

for example, involves one object in the focus of attention (the

figure) describing a motion path over the top of an object thatreceives less attention (the ground). Figure and ground in image-

schemas are termed trajectors and landmarks respectively. Trajec-

tors take the focus of attention at a cognitive level, and their reali-

zation in linguistic form is an example of the cognitive linguistic

principle that sees cognitive and perceptual processes as the basis

for language.

See also: cognitive linguistics; foregrounding; landmark

Further reading: Langacker 1987–91, Ungerer and Schmid 1996.

TRANSCRIPTION

A representation of speech in written form, using conventional

symbols. Since conventional writing systems are almost never ade-quate for representing pronunciation in a fully explicit and con-

sistent manner, phoneticians and linguists have found it necessary

to invent their own systems of symbols for transcribing speech

sounds, individual words and connected speech.

There are two types of transcription that indicate pronunciation,

and the difference is very important, though both types commonly

TRANSCRIPTION

301

use the International Phonetic Alphabet. In a phonetic transcription,the object is to record the physical and objectively real speech

sounds in as much detail as is required for the current purpose, and

two transcriptions of the same utterance may differ in the amountof detail they include. A phonetic transcription is always enclosed

in square brackets. So, for example, if we take a typical pronuncia-

tion of the English word please, the phonetic transcription [pliz]

presents only the bare minimum of phonetic information. For many

purposes, we would prefer the more detailed transcription [phli:z],which explicitly records both the aspiration (puff of breath) fol-

lowing the [p] and the length of the vowel [i:]. A still more detailed

version would be [phl8i:z], which notes that the [l] is voiceless,because of the presence of the aspiration. If the speaker is Amer-

ican, we might write [ph�i:z], where the symbol [�] explicitly indicates

the typical American ‘dark l’ (an [l] pronounced with the back of

the tongue raised). The amount of phonetic detail that might be

included in a phonetic transcription is almost unlimited, but we

normally content ourselves with recording only the information

that seems relevant to our purpose. The more detail we include, the

narrower is the transcription; the less detail we include, the broaderis the transcription.

It is important to note that a phonetic transcription records

physical reality. Hence a trained phonetician can successfully tran-

scribe any piece of speech, even one in a language unknown to her

(or him). She does not need to know anything about the language

she is listening to; she only needs to be able to hear the speech

sounds being produced and to identify them.

A phonemic transcription is very different. In a phonemic tran-scription, we do not transcribe any physical speech sounds at all.

Instead, we transcribe the phonemes of the language we are listen-

ing to, the basic sound units of that language. But this is only pos-

sible after we have first carried out a complete phonological

analysis of the language and decided what phonemes exist, which

speech sounds belong to which phonemes, and what symbols we

will use to represent the phonemes. It is therefore not possible to

produce a phonemic transcription of a totally unfamiliar language.A phonemic transcription is always enclosed in phoneme slashes.

Since we already have a complete phonemic analysis of English, we

are therefore able to provide a phonemic transcription of please, as

soon as we agree on the phoneme symbols to be used. Naturally

enough, we all prefer /p/ for the first consonant phoneme, /l/ for the

second consonant phoneme, and /z/ for the last one, but there is

TRANSCRIPTION

302

some disagreement about the vowel: some linguists prefer to repre-

sent the vowel phoneme as /i/, but others prefer the lengthened

form /i:/, and so please is phonemically either /pliz/ or /pli:z/,depending on which symbol we prefer for the vowel phoneme. (Wemust be consistent, of course.)

Transcripts are also used in sociolinguistics, discourse analysis and

computational linguistics to indicate the written record of speech.

Here, the transcript is often written in conventional spelling, per-

haps simply with the obvious features of writing omitted or used

differently: for example, capital letters for names and the beginnings

of sentences are not used, full-stops and dashes indicate short

pauses rather than grammatical markers, square brackets are usedto show where speakers overlap, and so on. The form of the tran-

script is a matter of the degree of detail required in the analysis: if

you are interested in syntactic expressions, for example, there is no

need for the detail of a phonetic transcription.

See also: International Phonetic Alphabet; phonetics; phonology

Further reading: Ball and Rahilly 1999; Ladefoged 1993.

TRANSFORMATIONAL GRAMMAR

A particular type of generative grammar. In the 1950s, Noam

Chomsky introduced into linguistics the notion of a generative

grammar, which has proved to be very influential. Now there are

very many different types of generative grammar which can beconceived of, and Chomsky himself defined and discussed several

quite different types in his early work. But, from the beginning, he

himself favoured a particular type, to which he gave the name

transformational grammar, or TG; TG has sometimes also been

called transformational generative grammar.Most types of generative grammar in which anybody has ever

been interested can be usefully viewed as working like this: starting

with nothing, the rules of the grammar build up the structure of asentence piece by piece, adding something at each step, until the

sentence structure is complete. Crucially, once something has been

added to a sentence structure, it must remain: it cannot be changed,

deleted or moved to a different location.

TG is hugely different. In TG, the structure of a sentence is first

built up in the manner just described, using only context-free rules,


303

which are a simple type of rule widely used in other types of gen-

erative grammar. The structure which results is called the deep

structure of the sentence. But, after this, some further rules apply.

These rules are called transformations, and they are different innature. Transformations have the power to change the structure

which is already present in a number of ways: not only can they add

new material to the structure (though only in the early versions),

but they can also change material which is already present in var-

ious ways, they can move material to a different location, and they

can even delete material from the structure altogether. When all the

relevant transformations have finished applying, the resulting

structure is the surface structure of the sentence. Because of the vastpower of transformations, the surface structure may look extremely

different from the deep structure.

TG is thus a theory of grammar which holds that a sentence

typically has more than one level of structure. Apart from the

structure which it obviously has on the surface, it also has an

abstract underlying structure (the deep structure) which may be

substantially different. The point of all this, in Chomsky’s view, is

that certain important generalizations about the structures of thesentences in a language may be stated far more easily in terms of

abstract deep structures than otherwise; in addition, the meaning of

a sentence can often be determined much more straightforwardly

from its deep structure.

TG has developed through a number of versions, each succeeding

the other. In his 1957 book Syntactic Structures, Chomsky provided

only a partial sketch of a very simple type of transformational

grammar. This proved to be inadequate, and, in his 1965 bookAspects of the Theory of Syntax, Chomsky proposed a very differ-

ent, and much more complete, version. This version is variously

known as the Aspects model or as the Standard Theory. All text-

books of TG published before 1980 (and a few of those published

more recently) present what is essentially the Standard Theory,

sometimes with a few additions from later work.

Around 1968 the Standard Theory came under attack from a

group of younger linguists who hoped to equate deep structure,previously a purely syntactic level of representation, with the seman-

tic structure of a sentence (its meaning). This programme, called

Generative Semantics, led to the positing of ever more abstract

underlying structures for sentences; it proved unworkable, and it

finally collapsed. Around the same time, two mathematical linguists

demonstrated that standard TG was so enormously powerful that it


304

could, in principle, describe anything which could be described at

all – a potentially catastrophic result, since the whole point of a

theory of grammar is to tell us what is possible in languages and

what is not possible. Yet these Peters-Ritchie results suggested thatTG was placing no constraints at all on what the grammar of a

human language could be like.

Chomsky responded to all this in the early 1970s by introducing

a number of changes to his framework; the result became known as

the Extended Standard Theory, or EST. By the late 1970s further

changes had led to a radically different version dubbed the RevisedExtended Standard Theory, or REST. Among the major innovations

of the REST were the introduction of traces, invisible flags markingthe former positions of elements which had been moved, a reduc-

tion in the number of distinct transformations from dozens to just

two, and a switch of attention away from the transformations

themselves to the constraints which applied to them.

But Chomsky continued to develop his ideas, and in 1981 he

published Lectures on Government and Binding; this book swept

away much of the apparatus of the earlier transformational theories

in favour of a dramatically different, and far more complex,approach called Government-and-Binding Theory, or GB. GB retains

exactly one transformation, and, in spite of the obvious continuity

between the new framework and its predecessors, the name ‘trans-

formational grammar’ is not usually applied to GB or to its even

more recent successor, the minimalist programme. Hence, for pur-

poses of linguistic research, transformational grammar may now be

regarded as dead, though its influence has been enormous, and its

successors are maximally prominent.

See also: derivation; generative grammar; Government-and-Binding Theory

Further reading: Antony and Hornstein 2003; Cook and Newson 1996;Grinder and Elgin 1973; Lyons 1991; Malmkjær 2004.

TRANSITIVITY

The manner in which a verb is related to the noun phrases in its

clause. Grammarians have been aware of differences in transitivity

since ancient times, though in the twentieth century we have

become aware that transitivity is a more complex matter than had

previously been thought.

TRANSITIVITY

305

We may begin by noting a fundamental difference between two

types of clause (or sentence). In a prototypical intransitive con-

struction, the (intransitive) verb has a subject but no object: Susie

smiled; Susie is vacationing in Bermuda. Here it would make nosense to ask ‘What did Susie smile?’ or ‘What is Susie vacationing?’

In a prototypical transitive construction, the (transitive) verb has

both a subject and an object; the subject represents an agent insti-

gating the action, and the object represents a patient affected by the

action: Susie slapped Dave; Susie is ironing a skirt.

However, in English and in other languages, these two construc-

tions are also used to express states of affairs which are less than

prototypical. For example, Susie smokes and Susie is eating areclearly intransitive, and yet here it does make sense to ask ‘What

does Susie smoke?’ or ‘What is Susie eating?’ In some sense, the

verbs smoke and eat are really transitive, but we do not bother here

to identify the object (what is smoked or eaten), because it is

obvious or unimportant. Such an apparently intransitive construc-

tion is sometimes called an absolute transitive construction.

Consider the following sentences, all of which are transitive in

form: Susie bought a car; Susie speaks French; Susie understands ourproblem; Susie weighs 110 pounds. These illustrate steadily decreas-

ing levels of prototypical transitivity: Susie is less and less of an

agent, and the object is less and less affected by the action – indeed,

the last two don’t really involve any action at all. In short, the

world provides a very wide range of possible relations between

entities, but English, like many other languages, provides only two

grammatical constructions, and every possibility must be squeezed

into one or the other of the two constructions.Unlike some other languages, English exhibits a special type of

transitive construction in which the verb takes two objects. Examples

of this ditransitive construction are Susie gave Natalie a kiss and Susie

showed me her new skirt. And, like many other languages, English

also has a special type of intransitive construction in which the verb

is followed by a noun phrase which does not behave like an object,

as in Susie is our regional manager and Susie has become a mother.

Within Systemic Linguistics, the notion of transitivity has beengreatly extended and generalized; here the term is understood as

denoting the kind of activity or process expressed by a sentence, the

number of participants involved and the manner in which they are

involved.

See also: verb

TRANSITIVITY

306

Further reading: Collins Cobuild 1990; Halliday 2004; Kilby 1984;Thompson 2004.

TREE

A particular type of graphical representation of the structure of

a sentence. Most linguists believe that the structure of any sen-

tence is typically a constituent structure, in which the sentence

consists of some pieces, and each piece consists of some smaller

pieces, and so on, down to the smallest pieces; moreover, each

piece belongs to some particular syntactic category. This kind ofstructure can be vividly illustrated by a graphical device called a

tree.Consider the sentence The little girl washed her doll. The tree

below illustrates its structure.

Every branch of the tree represents a single constituent of the

sentence, and every constituent, or node, is labelled with a node

TREE

307

label explaining to which syntactic category it belongs. These

standard abbreviations have the following meanings: S = Sentence;

NP = Noun Phrase; VP = Verb Phrase; Det = Determiner; N0 =N-bar; V = Verb; AP = Adjective Phrase; A = Adjective; N =Noun. A tree of this sort may also be called a tree structure, a treediagram or a phrase marker.

The precise nature of the tree drawn depends to some extent on

the particular grammatical framework being used. The tree shown

above would be more or less the tree preferred by most linguists,

but the proponents of certain frameworks, most notably the

Chomskyan framework called Government-and-Binding Theory,

would draw a much more elaborate tree involving a number ofabstract nodes containing no overt material; the Chomskyans

would also regard the tree as strictly representing only the surface

structure of the sentence, which might have a significantly different

deep structure, also representable as a tree.

See also: constituent structure; phrase-structure grammar; surface struc-ture

Further reading: Kroeger 2004.

TROPE

In rhetoric, the tropes were figures of speech that were supposed to

express certain emotional and persuasive effects. The main tropesare metaphor, analogy, simile, metonymy, synecdoche, litotes, hyper-bole, irony and so on. Traditionally these were largely defined in

terms of their stylistic realization rather than their conceptual con-

tent. The difference between metaphor and simile, for example, was

simply seen as the difference between an assertion (A is B) and a

comparison (A is like B). However, it has been argued in modern

cognitive linguistics that most of these are surface realizations of an

underlying conceptual metaphor, in the sense that they all replacethe lexical expression of the literal referent with an alternative.

Conceptually, the only difference between a metaphor and a simile

is in the degree of stylistic commitment to the meaning. Conceptual

metaphor is presented as a mapping between a source domain and a

target domain. The only difference is in the recent work which has

seen a metonymy as being a mapping within a single domain. The

TROPE

308

fields of stylistics and cognitive poetics have again become interested

in the linguistic texture of tropes.

See also: metaphor; rhetoric; stylistics

Further reading: Stockwell 2002b; Goatly 1987.

TURN-TAKING

That aspect of conversational structure by which the identity of the

speaker changes from time to time. In our idealized notion of a

conversation, we expect the floor to pass from one individual to

another in an orderly manner: one person speaks while the othersremain silent, and then the speaker falls silent and another person

takes the floor. What is interesting about this is that there appear to

be clear rules determining when and how the floor is handed over

from one person to another: if there were not, a conversation would

be merely a noisy jumble of several people trying to speak at once.

It was the American sociologist Harvey Sacks and his colleagues

who first drew attention, in the 1970s, to the importance of turn-

taking and the rules governing it. However, these rules are by nomeans easy to discover, and sociolinguists have in recent years

devoted a good deal of study to trying to elucidate them. Moreover,

it seems clear that very young children do not understand the con-

cept of turn-taking, and specialists in language acquisition are

therefore curious to find out how an understanding of turn-taking

is acquired. Similarly, the operation of turn-taking and the different

ways in which keeping a turn, giving up a turn and interrupting

indicate power dynamics have generated interest from discourseanalysts and sociolinguists, especially with reference to the way that

men, women and mixed groups conduct conversations.

See also: adjacency pair; communicative competence; ethnography ofcommunication; sex differences in language

Further reading: Coates 1996; Duranti 1997.

TYPOLOGY

The classification of languages according to their structural fea-

tures. One way of classifying languages is according to the genetic

TYPOLOGY

309

relationships among them; that is, according to their historical

origin. However, we can also classify languages according to the

kinds of structures they exhibit. Such a classification is typological,and a typological classification is, in principle, entirely independentof the histories of the languages involved – though languages which

are genetically rather closely related, naturally, are often also very

similar typologically.

In principle, we might pick on virtually any structural feature and

use it as the basis of a classification. For example, we could divide

languages into those in which the word for a canine animal is [dOg]and those in which it isn’t. (The first group here would contain

exactly two known languages: English and the Australian languageMbabaram.) But such a classification would be pointless, since it

wouldn’t lead anywhere.

The only typological classifications which are of interest are those

which are fruitful. By this we mean that the languages in each

category should turn out to have other features in common, fea-

tures which were not used to set up the classification in the first

place.

For example, we might classify languages into those which havetones (like Chinese and Yoruba) and those which don’t (like English

and Japanese). But this isn’t very fruitful, since neither the tone

languages nor the non-tone languages turn out to have much of

anything else in common, apart perhaps from an interesting geo-

graphical distribution, with tone languages concentrated in east

Asia and west Africa.

Another possibility is to classify languages according to their

morphology, or word-structure. In the early nineteenth century,Wilhelm von Humboldt tried this. He classified languages into three

types: isolating (a word typically consists only of a single mor-

pheme, as in Vietnamese or classical Chinese), agglutinating (a word

typically consists of a neat linear sequence of morphemes, all

clearly recognizable, as in Turkish or Swahili), and inflecting (a

word typically consists of several morphemes which are tangled up

together in a messy way and cannot easily be separated, as in Latin

or Russian). While famous, and possibly descriptively useful, thisclassification too has failed to be very fruitful, though it was later

developed further by Edward Sapir, who proposed several numer-

ical indices for characterizing word structure in some detail.

But the most celebrated and fruitful of all typological classifica-

tions has proved to be one in terms of basic word order. Proposed

by Joseph Greenberg in 1963 and more recently developed by John

TYPOLOGY

310

Hawkins and others, word-order typology has revealed a number of

striking and previously unsuspected correlations. For example, a

language with SOV order is highly likely to have modifiers that

precede their head nouns, auxiliaries that follow their main verbs,postpositions instead of prepositions, and a rich case system for

nouns. AVSO language, in contrast, usually has modifiers that fol-

lows their nouns, auxiliaries that precede their verbs, prepositions,

and no cases.

See also: basic word order; morphology; universal

Further reading: Comrie 1989; Croft 2002; Crystal 1997; Hawkins 1983;Horne 1966; Nevalainen et al. 2006; Whaley 1997.

UNIVERSAL

A statement which is true of all languages. Investigators of the past,

such as the eighteenth-century German philosopher Wilhelm von

Humboldt, were sometimes interested in searching for universalproperties of language. But the modern interest in the subject really

begins with the American linguist Joseph Greenberg in the 1960s.

Since Greenberg’s pioneering work, which led to a major project

based at Stanford University, a number of other linguists have

thrown themselves into the search for universals of language.

A separate strand of investigation was opened up around the

same time by the American linguist Noam Chomsky, who proposed

a search for universal grammar. As a result, we now distinguishbetween formal universals – universal constraints upon the form the

grammar of any language can take, and usually stated at a high

level of abstraction within some particular theory of grammar – and

substantive universals – statements about the linguistic objects which

can or cannot be present in languages and about their behaviour.

Here are a few substantive universals which have so far stood up

well: every language distinguishes nouns and verbs; every language

distinguishes three persons; every language has at least three vowels(the Caucasian language Kabardian is a possible exception to this

last one, with perhaps only two vowels).

A near-universal is a statement which has only a very few excep-

tions. Here is an example: every language has at least one nasal

consonant. This is almost true, but there are a couple of exceptions,

which therefore become intriguing.

UNIVERSAL

311

A particularly rewarding approach has been the identification of

implicational universals, which have the form ‘if a language has

property P, then it must also have property Q’. Pioneered by

Greenberg, this approach has been particularly developed by JohnHawkins. An example: if a language has front rounded vowels, then

it also has back rounded vowels.

The search for universals may sometimes lead to a typology. If we

contemplate a conceivably universal property P, and we then find

that a number of languages lack P, we may be able to set up a

typological classification of P-languages versus non-P-languages. Of

course, like any typology, this one will only be interesting if it turns

out that the languages in each group fairly consistently have other

properties in common besides P or not-P.

See also: typology; universal grammar

Further reading: Comrie 1989; Croft 1990; Crystal 1997; Greenberg1963; Hawkins 1983.

UNIVERSAL GRAMMAR

The grammatical properties shared by all human languages. When

the American linguist Noam Chomsky introduced his innateness

hypothesis in the 1960s, he was arguing that important parts of the

structure of human languages are built into our brains at birth.

Since about 1980, Chomsky has been elaborating his position and

arguing that certain fundamental principles for constructing sen-

tences can be found in all languages and must be part of ourgenetic endowment, present from birth. These principles he collec-

tively terms universal grammar.Naturally, Chomsky attempts to formulate these principles within his

own theoretical framework, although that framework has changed

dramatically over the years. The principles are necessarily rather

abstract, having to do with things like how far apart two linguistic

elements can be within a sentence when they are linked in some way.

Chomsky and his followers are convinced that such principlesgenuinely exist, and much of their work has been devoted to unco-

vering them. But this has proved to be difficult: proposed principles

are endlessly reformulated to cope with recalcitrant data, and a

number of apparent counterexamples are dealt with by appealing to

various theoretical gadgets which allow the principles to be violated

in certain circumstances. Critics are often deeply suspicious of these

UNIVERSAL GRAMMAR

312

seemingly ad-hoc manoeuvres, and some of them doubt whether

the principles of universal grammar are really there at all.

See also: Government-and-Binding Theory

Further reading: Antony and Hornstein 2003; Cook and Newson 1996.

USAGE

The collective speaking and writing habits of a particular group of

people, or a particular one of these habits. English, like any lan-

guage, is far from being completely uniform. In particular, certain

words, spellings, pronunciations, meanings and grammatical forms

are not commonly found everywhere, but are typical only of certainplaces, certain professions or certain styles. Consequently, a careful

description of English must provide this information, and good

dictionaries of English regularly do this. Dictionaries routinely label

certain forms with annotations such as (for geography) Australian,

Scots, chiefly American, (for technical terms in various professions)

mathematics, botany, nautical, grammar, and (for style) slang, taboo,

archaic, informal. Such usage labels indicate that a particular usage

is only appropriate in some particular context, and would be unex-pected or out of place elsewhere.

It should be stressed that most items marked with such usage

labels are no less a part of English than other items, and it is a

misunderstanding to conclude that, say, slang terms or swear words

are not ‘real English’. But archaic words are something of a special

case. These are words which have dropped out of use and no longer

form a part of the modern language, except perhaps in deliberately

archaic contexts like religious language, and dictionaries sometimesinclude them purely as a convenience to users who are trying to

read texts written centuries ago. For example, because of the

importance we attach to the works of Shakespeare, some dictionaries

make a point of including every single word and word-meaning

found in Shakespeare, even if these are found nowhere else at all.

The application of usage labels is a natural part of the approach

we call descriptivism: describing the facts of a language as they

appear to exist.

See also: descriptivism


USAGE

313

UTTERANCE

A particular piece of speech produced by a particular individual on

a particular occasion. In linguistics, a sentence is an abstract lin-guistic object forming one part of the total expressive resources of a

given language. When we speak, therefore, we do not strictly pro-

duce sentences: instead, we produce utterances. An utterance is a

single piece of speech marked off as a unit in some way; for example,

by pauses and intonation.

There is only one English sentence of the form What’s for dinner?

But, every time you say What’s for dinner?, you are producing a

different utterance. Each one of these utterances may differ notice-ably from others: it may be faster or slower, louder or softer; one

may be cheerful or eager, while another may be bored or suspicious,

and the particular social context in which it is uttered will vary. But

every one of these utterances corresponds to the same English sentence.

Moreover, an utterance need not correspond to a sentence at all.

Consider the following exchange. Mike: Where’s Susie? Sarah: In

the library.

Here Mike’s utterance corresponds to an English sentence, butSarah’s response does not: there is no English sentence of the form

*In the library (as the asterisk indicates). Of course, Sarah’s

response is perfectly normal: it’s just that not all of our utterances

correspond to sentences. Instead, some of them correspond only to

fragments of sentences.Even more dramatically, we frequently produce utterances which

are interrupted or broken off. Examples: Mike, would you get me . . .oh, never mind; I’ll get it myself; I was just looking for the . . . um . . .[tails off into silence]. These are still utterances, but obviously they

do not correspond directly to any sentences.

See also: sentence, speech

Further reading: Cutting 2002; Grundy 1995; Verschueren 1999.

VARIABLE

A key element in sociolinguistics, the notion of the variable is based

on the premise that a linguistic feature can have different realiza-

tions and that these are distributed in correspondence with a social

factor. The linguistic variable is a language feature (most usually a

UTTERANCE

314

phonetic feature of a particular accent, though lexical and gram-

matical variables have also been used) which has at least two dis-

tinguishable variants. The distribution of these variants can be

shown not to be random but to have a correlation with a socialvariable such as age, class, gender, geography, politics, education,

and so on. Mapping the nature and frequencies of these correla-

tions is a major explanatory method in variationist sociolinguistics.In statistical terms, the linguistic variable is the dependent variable,and the social variable is the independent variable.

For example, in the most famous small-scale study in socio-

linguistics, the American linguist William Labov explored the pro-

nunciation of the linguistic variable of rhoticity (pronunciation ofnon-prevocalic /r/) in the phrase ‘fourth floor’. By asking a question

to which he knew this was the answer, and then pretending to mis-

hear so that the informant repeated it, he was able to observe four

occurrences of /r/: word-medial non-prevocalic; word-final non-

prevocalic; emphatic word-medial non-prevocalic; and emphatic

word-final non-prevocalic. By asking these in three different New

York department stores of hierarchical status, he was able to

demonstrate a prestige value for rhoticity in New York accents.Many sociolinguistic studies have applied this basic principle since.

See also: accent; sociolinguistics

Further reading: Llamas et al. 2007; Stockwell 2002a; Wardhaugh 2005.

VARIATION

The existence of observable differences in the way a language is

used in a speech community. It is a commonplace observation that

a single language is not used in a totally homogeneous manner

within a single community. Stockbrokers do not speak like plum-

bers; men do not speak like women; older people do not speak like

younger people. Moreover, even the speech of a single person is not

homogeneous: you don’t speak in the same way when you’re chat-ting to friends in a bar and when you’re being interviewed for a job,

and even in a single context you might say telephone or I got chea-

ted one moment but phone or I got ripped off the next.

This variation was long regarded by most linguists as no more

than a collection of tiresome details getting in the way of good

descriptions, as something to be swept under the carpet and

VARIATION

315

forgotten about. But times have changed. In the 1960s, the socio-

linguists, led by the American William Labov, began to make var-

iation a central object of investigation, and the result has been a

revolution in linguistics: we now realize that variation, far frombeing peripheral and inconsequential, is a vital part of ordinary

linguistic behaviour.

To begin with, variation often shows strong correlations with

social variables like social class and sex, and the social stratification

of language is now a prominent feature of sociolinguistic investiga-

tions. Further, the introduction of the quantitative approach to lan-

guage description has revealed important patterns of linguistic

behaviour which were previously invisible. The concept of a socio-linguistic variable has become central to the description of speech.

A variable is some point of usage for which two or more competing

forms are available in a community, with speakers showing inter-

esting and significant differences in the frequency with which they

use one or another of these competing forms.

Furthermore, it has been discovered that variation is typically the

vehicle of language change, and as a result a satisfactory resolution

has finally been obtained for the Saussurean paradox.

See also: quantitative method; Saussurean paradox; social stratificationof language; sociolinguistics

Further reading: Chambers 2003; Chambers et al. 2002; Foulkes andDocherty 1999; Holmes 1992; Hudson 1996; Llamas et al. 2007.

VERB

The part of speech which includes words like go, see, understand and

seem. The class of verbs is universal: no language has ever been

discovered which lacked a distinct class of verbs. The most proto-

typical verbs denote actions performed by an agent, such as run,

sing, throw, hit and give. But many other items are verbs even

though they have less typical meanings, such as die, sleep, believe,understand, elapse, ensue, become, seem, have and be. The English

auxiliaries, like must and should, are also usually classed as verbs,

though a few linguists disagree here.

What unites the class of verbs is their grammatical behaviour.

For example, verbs in English and in many other languages are

marked for tense: Susie drinks brandy versus Susie drank brandy.

VERB

316

Verbs also frequently exhibit agreement, though English has only a

tiny amount of this, as in Susie smokes versus Susie and Janet smoke.

But the most central characteristic of a verb is the requirement

that it must be accompained by one or more noun phrases, itsarguments, in a grammatical sentence. For example, smile and

smoke take only one argument (Susie smiled; Susie smokes); buy and

kiss require two (Susie bought a car; Susie kissed Natalie); and give

and show require three (Susie gave me this book; Susie showed Mike

her new car). A very few verbs, though, are unusual, such as rain

and seem: It’s raining; It seems that Susie is away.

These differences in grammatical behaviour illustrate sub-

categorization, and they involve differences in transitivity.

See also: subcategorization; tense; transitivity

Further reading: Greenbaum and Quirk 1990; Hurford 1994; Jackson2002.

VERB PHRASE

A unit of sentence structure consisting of a verb and the other ele-

ments closely linked to it. Traditional grammarians divided a typi-

cal sentence into two parts: a subject and a predicate. In our

modern elaboration of this, the syntactic category which acts as a

subject is a noun phrase, while the category acting as a predicate is a

verb phrase. A verb phrase always contains a verb, and it usually

contains some other material closely linked to that verb.Here are some examples, with the verb phrases serving as pre-

dicates bracketed: Susie [smokes]; Susie’s sister [has a new girl-

friend]; The rest of us [strolled down to the pub for a drink].

A verb phrase need not serve as a predicate. In the following

examples, the bracketed verb phrases are doing something different:

[Smoking forty cigarettes a day] is bad for your health; [Needing some

legal advice], Susie rang her lawyer; Susie has decided [to buy a new car].

When a sentence contains an auxiliary, linguists differ in theiranalyses. For most linguists, the sentence Susie has finished her

dinner contains two verb phrases, a smaller one inside a larger one,

as follows: Susie [has [finished her dinner]]. Others, though, would

recognize only one verb phrase: either Susie [has finished her dinner]

or Susie has [finished her dinner], with the auxiliary excluded from

the verb phrase in this last analysis.

VERB PHRASE

317

With multiple auxiliaries, things become more complicated. Most

linguists would analyse Susie [has [been [smoking a lot]]] as con-

taining three verb phrases, as shown. The first dissenting view

would again recognize only one, as in Susie [has been smoking a lot],while the other has a serious problem, possibly to be resolved as

Susie [has been] [smoking a lot].

See also: predicate; syntactic category; verb

Further reading: Gramley and Patzold 1992; Greenbaum and Quirk1990; Jackson 2002.

VERNACULAR

The ordinary, everyday speech of a particular community, some-

times informally called its mother tongue. The term vernacular is

most commonly contrasted with standard language. The vernacular

speech of a particular community is the ordinary speech used by

the people in a particular community, such as Chicago, Liverpoolor Sydney, when this is noticeably different from the standard form

of the language.

Especially where European languages were concerned, the lin-

guists of the past normally concentrated on the standard forms of

languages. Non-standard vernacular forms were silently ignored,

except only in the study of regional dialects, for which the speech of

elderly rural speakers was considered most appropriate; at the same

time, the speech of younger speakers or of urban speakers wassimilarly ignored.

Interest in vernacular forms developed only slowly during the

twentieth century, but it became increasingly prominent with the

rise of sociolinguistics in the 1960s. Today, there is intense interest in

vernacular forms of speech, which are now seen as every bit as

worthy of study as standard varieties of languages. Most prominent

so far is the study of vernacular forms of English, for which the

British sociolinguists Jim and Lesley Milroy have recently coinedthe slightly facetious term real English, partly in recognition of the

observation that certain forms and usages absent from standard

English, such as he don’t, are extremely widespread in vernacular

forms of English almost everywhere.


VERNACULAR

318

Further reading: Cheshire and Stein 1997; Milroy and Milroy 1993,1997.

VOCAL TRACT

The part of our body through which air passes during speech. The

vocal tract runs from the lungs up through the trachea (or wind-

pipe), through the pharynx (the space at the back of the mouth),

and there it divides into the oral cavity (the space inside the mouth)

and the nasal cavity; it reaches the outside world at the lips and at

the nostrils. As a general rule, during speech, air flows upwards andoutwards through the vocal tract, though in certain less usual air-

stream mechanisms something different happens.

In the throat, the vocal tract passes through a complex structure

called the larynx; within the larynx is an opening called the glottis,which can be opened wide to allow air to pass through freely, closed

tight to block the flow of air, or closed loosely. In this last case, two

bands of tissue called the vocal folds – between which the glottis is

located – undergo vibration, and this produces voicing.Between the oral cavity and the nasal cavity is a hinged flap of

tissue called the velum (or ‘soft palate’); when the velum is raised,

the nasal cavity is closed off, and no air can flow through it, but,

when the velum is lowered, air can flow out through the nose.

Within the oral cavity, the size and shape of the vocal tract can

be greatly varied, by raising or lowering the jaw, by moving the

tongue around, and by altering the position of the lips. Whenever

these movements are such as to greatly obstruct the flow of air, theresulting sound is a consonant; when the obstruction is minimal, the

result is a vowel.

See also: airstream mechanism; phonetics; voicing

Further reading: Ashby 1995; Ball and Rahilly 1999; Clark and Yallop1995; Collins and Mees 2003; Denes and Pinson 1993.

VOICE

The grammatical category governing the way the subject of a sen-

tence is related to the action of the verb. English has only a two-

way distinction of voice. In the active voice, the subject of the

VOICE

319

sentence is typically the entity performing the action, as in Tamer-

lane imprisoned Sultan Bayezit. In the passive voice, the subject is

instead the entity undergoing the action, as in Sultan Bayezit was

imprisoned by Tamerlane, or in Sultan Bayezit was imprisoned. Thefirst is called the long passive, or passive-with-agent; the second is

the short passive.In English, the active voice is unmarked (see markedness): it is

grammatically simpler and far more frequent in speech. The passive

voice is marked, and it is most typically used either to make the

entity undergoing the action the centre of attention, or to remove

the entity performing the action (the agent) from the centre of

attention, and possibly to remove it from the sentence altogether.For example, we write Uranium was discovered in 1789, but its

importance was not recognized before the middle of the twentieth

century. Here the dates are far more important than the name of

the discoverer, and identifying the people who failed to recognize

the importance of uranium would be pointless, since these people

included everybody.

Not all languages have a contrast between active and passive

voices. But some other languages have additional voices, allowingthe subject of the sentence to be not only the performer of the

action or the recipient of the action, but also the instrument with

which the action is performed, or the place in which the action is

done.

See also: verb

Further reading: Greenbaum and Quirk 1990; Hurford 1994; Jackson2002; Klaiman 1991; Palmer 1971.

VOICING

Vibration of the vocal folds. The vocal folds (or vocal cords) are

two movable masses of tissue in the larynx (the voice box). They

can be brought close together so that they vibrate all along theirlength as air flows up from the lungs and into the pharynx, mouth

and nose. This vibration is voicing, and any speech sound which

is produced with such vibration is voiced. Examples of voiced

sounds are [a], [w], [n], [l], [z], [b] and [dZ]. If you put your fingers

to your throat while pronouncing one of these, you can feel the

vibration.

VOICING

320

If the vocal folds are moved farther apart, they cannot vibrate. A

speech sound produced without vibration is voiceless. Examples of

voiceless sounds are [f], [s], [p], [k], [h] and [tS]. The glottal stop, [?],is also voiceless, but for a different reason: here the vocal folds arepressed so tightly together that no air can flow through them and

hence no vibration occurs.

The vocal folds can also behave in more complicated ways, pro-

ducing several complex phonation types in addition to voicing and

voicelessness.

Note that the term voicing is also applied to something else: to a

change in pronunciation in which a formerly voiceless sound

becomes voiced. For example, the Spanish phoneme /s/ is normallya voiceless sound [s], but in certain circumstances it becomes a

voiced sound [z], as in mismo ‘same’ (in standard European pro-

nunciation, however, not all speakers do this).

See also: consonant; phonation type

Further reading: Ball and Rahilly 1999; Catford 1988; Collins and Mees2003; Laver 1994.

VOWEL

A speech sound produced with no obstruction of the airstream.

From the point of view of phonetics, speech sounds may be con-

veniently divided into two types: those which are produced with a

substantial obstruction of the airstream and those which are not.The first are consonants, the second vowels.

If you pronounce a few vowels, such as [a], [i] and [u], you will

find that the flow of air through your vocal tract is nowhere impe-

ded. What makes one vowel sound different from another is the size

and shape of the space within your mouth: the jaw is higher or

lower, the tongue is higher or lower, either the front or the back of

the tongue may be raised, and the lips may be rounded or spread.

These variations affect the way in which the air in the mouth reso-nates, and they are responsible for the different qualities of the

vowels.

Certain speech sounds which are strictly vowels are commonly

treated as consonants, for a reason to be explained below. For

example, the English y-sound, as in yes and yard, whose phonetic

symbol is [j], is really nothing but a very brief version of the vowel

VOWEL

321

[i], as in see, and the [w] of weed and war is likewise only a brief

version of the vowel [u], as in moon. Moreover, for most (not all)

English-speakers, the /r/ at the beginning of red is really only a

slightly odd vowel pronounced with some part of the tongue raised.Yet all of these are generally classed as consonants.

The reason for this is that there is a second way of defining

vowels and consonants, one which gives different results from the

first. This second approach is the one preferred in phonology, in

which we are more concerned with the way sounds behave than

with their phonetic nature. And the point is that the sounds just

mentioned, even though they are phonetically vowels, behave like

consonants in English.A syllable always contains a peak of sonority, a part which is

louder and more sonorous than the rest, and in most languages

(including English), this peak must be a vowel. In the majority of

syllables, this vowel is preceded and/or followed by some number of

consonants. Usually these consonants are consonants by any

definition, as in key /ki:/, eat /i:t/, sat /sæt/, slap /slæp/ and fleeced

/fli:st/. However, the sounds /j/, /w/ and /r/ pattern like consonants,

not like vowels, as in yes /jes/, wet /wet/ and red /red/, and conse-quently they are classed as consonants, in spite of their phonetic

nature.

When you come across the word vowel, therefore, you must check

to see if it is being used in the phonetic sense or in the phonological

sense.

See also: consonant; speech sound; syllable

Further reading: Ball and Rahilly 1999; Collins and Mees 2003; Lade-foged 1993; Ladefoged and Maddieson 1996; Laver 1994.

WELL-FORMEDNESS

The status of a linguistic form in a language which conforms to the

rules of that language. Well-formedness is a central concept in theanalytical study of language, and it can be recognized at several

different levels of analysis.

At the level of phonology, it is clear that brick is a well-formed

English word, consisting of a permissible sequence of English con-

sonants and vowels. So is blick, even though no such word happens

to exist: this word could exist, since it obeys all the rules. But bnick

WELL-FORMEDNESS

322

is not well formed, since the rules of English do not permit a word

to begin with the sequence /bn-/. Even a real word may be ill

formed: the French word genre is well formed in French, but not in

English, and English-speakers must either struggle to give it aFrench-style pronunciation, or simply change the pronunciation to

conform to the rules of English, producing something that sounds

like jonra or jonner. Even sphere is arguably not well formed, since

English does not normally allow a word to begin with the sequence

/sf-/: nobody would name a new detergent Sfizz.

At the level of morphology, words like unhappiness and existenti-

alism are well formed, since they obey all the rules for combining

morphemes into words, but things like happy-un-ness and exist-al-

ism-ence are not, since they violate these same rules. Even though

they follow the regular rules for making plurals, the forms womans and

childs are not well formed, since woman and child happen to be subject

to special rules that require their plurals to be women and children.

At the level of syntax, There’s a spider on the bed is well formed,

but Bed the on spider a there’s is anything but well formed, since it

violates the rules of English word-order. Similarly, Susie has written

a letter is well formed, but Susie has writing a letter is not, since itviolates a rule of English syntax. Syntactic well-formedness is often

called grammaticality.

At the level of semantics, there are also rules for combining

meanings. For example, the phrase a tall woman is semantically well

formed, while a female woman is somewhat anomalous, since the

meaning of female is already present in the meaning of woman, and

a valid woman is virtually impossible to interpret sensibly. Jezebel

killed Ahab, but Ahab didn’t die is semantically anomalous, assum-ing only one Ahab is involved, since Ahab died is an integral part of

the meaning of killed Ahab. Making these make sense often involves

imagining different possible worlds in which physical laws or social

attitudes are at odds with our own.

At the level of pragmatics and discourse, some utterances and

exchanges are well formed while others are not. In reply to the

question Is Susie coming to Mike’s party on Saturday?, the response

I don’t think so is obviously well formed, and the response Natalie

wants to go to a concert might be well formed in certain circum-

stances (for example, if Natalie is Susie’s girlfriend), but the

response Susie used to be that little girl in the pizza commercial

would probably not be well formed in any circumstances.

Anything which is not well formed is ill-formed, or sometimes

anomalous. A very great deal of work in linguistics consists of

WELL-FORMEDNESS

323

identifying the rules governing well-formedness at all levels of

analysis.

See also: rule; structure


WORD

A linguistic unit typically larger than a morpheme but smaller than

a phrase. The term word might seem familiar and straightforwardenough, but in fact words can be defined in at least four different

ways, and these ways are not equivalent at all.

� An orthographic word is something written with white spaces at

both ends but no white space in the middle. Orthographic words

are of minimal linguistic interest.

� A phonological word is something pronounced as a single unit.

� A lexical item, or lexeme, is a dictionary word, an item whichyou would expect to find having its own entry in a dictionary.

� A grammatical word-form (GWF) (or morphosyntactic word) is

any one of the several forms which a lexical item may assume for

grammatical purposes.

Let’s look at some examples. The item ice cream is two ortho-

graphic words, but a single phonological word (it’s pronounced as a

unit), a single lexical item (it’s entered in the dictionary), and asingle GWF (indeed, it hardly has another form, unless you think

the plural ices cream is good English).

The singular dog and the plural dogs are each a single ortho-

graphic word, a single phonological word, and a single GWF, but

they both represent the same lexical item (they would only get one

entry in the dictionary). The same is true of take, takes, took, taken

and taking: five orthographic words, five phonological words, five

GWFs (at least), but only one lexical item. The two lexical itemshere would be entered in the dictionary as dog and take; these are

the citation forms, or the lemma form, of these lexical items, the

forms we use in naming them or talking about them.

The contraction hasn’t is a single orthographic word and a single

phonological word, but it’s two lexical items (have and not), and

two GWFs (has and not). The phrasal verb make up (as in She

WORD

324

made up her face) is two orthographic words, two phonological

words, but only one lexical item (because of its unpredictable

meaning, it must be entered separately in the dictionary). And it

has several GWFs: make up, makes up, made up, making up. Thevery different sense of make up illustrated by She made up a story

would be regarded by most linguists as a different lexical item from

the preceding one (a separate dictionary entry is required), but this

lexical item exhibits the same orthographic, phonological and

grammatical forms as the first.

Consequently, when you are talking about words, it is essential to

specify exactly which sense you have in mind, and it may be pre-

ferable to use one of the more specific labels.The study of words is lexicology.

See also: lemma; part of speech

Further reading: Hoey 2005; Jackson 2000; Katamba 1994; Matthews1991.

WORD-FORMATION

The process of constructing new words from existing materials.

There are many ways of constructing new words, and English uses

almost all of them.

In compounding, two (or more) existing words are simply com-

bined. There are several different patterns available for doing this,

as in blackboard, redneck, overthrow, olive green, scarecrow andforget-me-not. Sometimes an affix is also present, as in blue-eyed

and flat-earther.

In derivation, affixes are added to an existing word, as in pre-

history, rewrite, unsafe, washable, prohibition and finalize. Multiple

affixes are possible, as in misdirection, illegitimacy, transformational

and existentialism.

In clipping, a piece is removed from a longer word, as with bra,

gym, flu, phone and cello. In blending, pieces of two words arecombined, as in smog (smoke plus fog). In back-formation, a new

word is extracted from a longer word that appears to contain an

affix but historically does not, as with pea from earlier pease, edit

from earlier editor, and burgle from the earlier burglar.

Reanalysis involves changing the structure assigned to a word

and extracting a piece that formerly was not a part of it, as when

WORD-FORMATION

325

hamburger (Hamburg +-er) was reanalysed as ham +-burger, yield-

ing the -burger now used in cheeseburger and vegeburger.

Combining the initial letters of a phrase into a single word yields an

acronym if the result can be pronounced as a word (as with laser, fromlight amplification by the stimulated emission of radiation), but an

initialism if it must be spelled out letter by letter, as with FBI and BBC.

Most of our technical terms are constructed by gluing together

Greek and Latin combining forms of appropriate meaning: so, for

example, the recently discovered creatures which flourish at or

above the temperature of boiling water have been named hyper-

thermophiles, from three Greek elements meaning ‘high-heat-lover’.

See also: affix; derivation; morphology

Further reading: Adams 1973; Bauer 1983; Katamba 1994.

WRITING SYSTEM

A conventional system for representing a language with permanentmarks. Though human beings have been able to speak for many

tens of thousands of years, writing systems were invented only a

little more than 5,000 years ago, in Asia Minor. True writing was

preceded by precursors which were adequate for recording only

certain types of information, such as taxes due and paid. In a true

writing system, however, any utterance of the language can be ade-

quately written down.

There are several conceivable ways of constructing a writingsystem. We might attempt to provide a separate symbol for every

different word-form of a language. This would mean, for example,

that separate symbols would have to be provided for all of drive,

drives, driving, drove, driven, driver, drivers, driveway, and so on. But

the number of symbols required would be astronomical, and every

new word or word-form entering the language would require a new

symbol, and so such a system is completely unworkable.

Or we could provide a separate symbol for each morpheme in thelanguage. Thus drive would have its own symbol, but drives would

be written with the symbol for drive plus the symbol for present-

tense -s, and driver with the symbol for drive plus that for -er, and

so on. This is called a logographic system (though it might better be

called a morphemographic system), and it actually works fairly well

with a language whose words show little or no variation in form for

WRITING SYSTEM

326

grammatical purposes. Chinese is such a language, and still today

Chinese is written in a logographic script – though naturally the

number of different characters required still runs into many thou-

sands. But note that Chinese does not provide a separate characterfor every word. There is a character for the word huo ‘fire’ and

another for the word che- ‘vehicle’, but the word for ‘train’, huoche-,

literally ‘fire-vehicle’, is simply written with a combination of these

two characters.

A third possibility is to provide a separate symbol for every dis-

tinct syllable in the language. Japanese does this. So, the Japanese

word ikura ‘how many?’ is written with one symbol for /i/, a second

for /ku/, and a third for /ra/, and similarly for every other word(except that Japanese uses Chinese characters to write many words).

This kind of writing system is a syllabary, and it really only works

well with a language which has only a small number of distinct

syllables – no more than a few dozen. When a language permits

very many different syllables, then either the number of symbols

required becomes huge, or the system must be defective in some way –

for example, a single symbol may be used to represent a number of

distinct but somewhat similar syllables. The Linear B syllabary usedto write Mycenaean Greek was like this: for example, the same

symbol was used to write to, tho, tos, thon, and other syllables.

Finally, we can provide a separate symbol for each phoneme in

the language – that is, for each distinctive consonant and vowel.

Such a writing system is an alphabet, and alphabetic writing, the

last type to be invented, is now by far the world’s most frequent

type of writing. The first alphabets, invented by speakers of Semitic

languages thousands of years ago, were in fact defective, in thatthey provided letters (as alphabetic symbols are called) only for

consonants; it was the Greeks who constructed the first complete

alphabet by adding letters for vowels. In an ideal alphabetic writing

system, every consonant and vowel in the language has its own

consistent letter, and the spelling of a word is completely pre-

dictable from its pronunciation, but in practice few alphabetic

scripts approach this ideal. In English, with its exceptionally com-

plex and irregular spelling system, the spelling of a word typicallyexhibits only a modest correlation with its pronunciation (note the

spellings of rite, write, right and wright, the verb lead and the name

of the metal lead, and curiosities like debt, knight, buy, pharaoh and

autumn), and our supposedly alphabetic system has become more

similar to the Chinese logographic script, in which the representation

of each simple word has to be learned as a unit.

WRITING SYSTEM

327

Mixed systems are possible. The enormously complex Japanese

writing system mostly uses Chinese characters to represent the

stems of words, but a Japanese syllabary to represent grammatical

words and grammatical endings, and it uses a second syllabary forvarious special purposes, such as writing words of foreign origin.

The ancient Egyptian hieroglyphic writing system was almost

numbingly complex, using a mixture of characters of various types

to provide clues about both pronunciation and meaning.

In principle, any language can be written in any writing system,

though not always with the same degree of success. Several central

Asian languages have, during the past century, been successively

written in the Arabic alphabet, in the Roman alphabet and in theCyrillic (Russian) alphabet, depending on which way the political

winds were blowing; Korean is sometimes written in Chinese char-

acters but at other times in the local Han’gul alphabet; Chinese and

Japanese are both sometimes written in the Roman alphabet for

special purposes, but both the Chinese and the Japanese have so far

resisted pressures to change over completely to the Roman alpha-

bet. Turkish, in contrast, abandoned the Arabic alphabet in the

1920s in favour of the Roman alphabet, at least officially.Note carefully that every writing system ever used represents an

attempt at recording utterances in a particular language. There has

never been a real writing system which attempted to represent

‘ideas’ or ‘thoughts’ directly, without the mediation of a particular

language, and suggestions to the contrary are ignorant and fantas-

tic. (A few philosophers have occasionally tried to invent such sys-

tems, but they don’t work.)

The study of writing systems is sometimes called graphology (notto be confused with the ‘psychological’ interpretation of hand-

writing, a form of charlatanism of the same name).

See also: orthography; punctuation; spelling

Further reading: Biber 2006; Coulmas 1996; Crystal 1997; Daniels andBright 1996; Fromkin and Rodman 1998; O’Grady et al. 1996; Robin-son 1995; Sampson 1985.

X-BAR

A system of syntactic description based on the notion that every

constituent has a head element. The X stands for any lexical head,

X-BAR

328

such as N (noun), V (verb) or P (preposition), and the ‘bar’ refers to

the notational symbol placed over the X-element to indicate the

other elements of the construction: N–or N2, which can read as ‘N-

bar’ and ‘N-double-bar’. The system asserts that every syntacticcategory is complex. Even nouns with no other element count as

noun-phrases by virtue of having a zero entry for the ‘bar’ element.

This forces the analyst to treat every construction as having a head.

X-bar theory consists of three basic syntactic assembly rules,

which allow an enormous number of permutations, since they are

recursive. First, an X-bar consists of an X (a head) plus any number

of complements. Second, an X-bar can consist of an X-bar and an

adjunct, in any sequence (this gives the recursive property). Last, anX-Phrase consists of an optional specifier and an X-bar, in any

sequence. To give an example involving the last of these rules, a

simple noun phrase like this fat book consists of a N-bar (fat book)

plus a specifier (this). The N-bar phrase itself consists of another

N-Bar (the head book with a zero bar element) plus a specifier (fat).

The system – especially when drawn as a tree – gives the rules for

constituent structure. The other rules, and all of them applied to

verb-phrases and prepositional phrases, can be used to generate thesyntactic sequences of, it is claimed, all the languages of the world.

Almost all current theories of grammar employ some variant on

this system.

See also: head; phrase-structure grammar; transformational grammar;tree

Further reading: Borsley 1999; Cook and Newson 1996.

X-BAR

329

Bibliography

Aarts, Bas and McMahon, April (eds) (2006) The Handbook of EnglishLinguistics, Oxford: Blackwell.

Adams, Valerie (1973) An Introduction to Modern English Word-Formation, London: Longman.

Adolphs, Svenja (2006) Introducing Electronic Text Analysis, London:Routledge.

Auer, Peter, Hinskens, F. and Kerswill, Paul (eds) (2005) DialectChange, Cambridge: Cambridge University Press.

Aitchison, Jean (1994) Words in the Mind (2nd edn), Oxford: Blackwell.—— (1996) The Seeds of Speech, Cambridge: Cambridge University Press.—— (1998) The Articulate Mammal (4th edn), London: Routledge.—— (2001) Language Change: Progress or Decay? (3rd edn), Cam-bridge: Cambridge University Press.

Akmajian, Adrian, Demers, Richard A., Farmer, Ann K. and Harnish,Robert M. (1995) Linguistics: An Introduction to Language andCommunication, Cambridge, MA: MIT Press.

Allan, Keith (1986) Linguistic Meaning (2 vols), London: Routledge.Allen, Donald E. and Guy, Rebecca F. (1974) Conversation Analysis:The Sociology of Talk, The Hague: Mouton.

Antony, Louise M. and Hornstein, Norbert (eds) (2003) Chomsky andHis Critics, Oxford: Blackwell.

Anttila, Raimo (1988) An Introduction to Historical and ComparativeLinguistics (2nd edn), London: Macmillan.

Archangeli, Diana and Langendoen, D. Terence (eds) (1997) OptimalityTheory: An Overview, Oxford: Blackwell.

Ashby, Patricia (1995) Speech Sounds, London: Routledge.Asher, R.E. and Simpson, J.M.Y. (eds) (1994) Encyclopedia of Lan-guage and Linguistics (10 vols), Oxford: Pergamon.

Auer, Peter (ed.) (1998) Code-Switching in Conversation: Language,Interaction and Identity, London: Routledge.

Austin, J.L. (1962)How to Do Things with Words, Oxford: Clarendon Press.Bach, Emmon (1974) Syntactic Theory, New York: Holt, Rinehart &Winston.

Baker, Marc (1996) The Polysynthesis Parameter, Oxford: Oxford Uni-versity Press.

Bakhtin, Mikhail (1968) Rabelais and his World (trans. HeleneIswolsky), Cambridge, MA: MIT Press.

—— (1984) Problems of Dostoevsky’s Poetics (ed. and trans. CarylEmerson, Problemy Tvorchestva Dostoyevskogo, 1929), Manchester:Manchester University Press.

330

Baldi, Philip (1983) An Introduction to the Indo-European Languages,Carbondale and Edwardsville, IL: Southern Illinois University Press.

Ball, Martin and Rahilly, Joan (1999) Phonetics: The Science of Speech,London: Arnold.

Barnbrook, Geoff (1996) Language and Computers: A Practical Intro-duction to the Computer Analysis of Language, Edinburgh: EdinburghUniversity Press.

Barss, Andrew (ed.) (2003) Anaphora: A Reference Guide, Oxford:Blackwell.

Bates, Elizabeth (1976) Language and Context: The Acquisition ofProgmatics, New York: Academic Press.

Bates, Elizabeth, Benigni, L., Bretherton, Inge, Camaioni, L. and Vol-terra, V. (1979) The Emergence of Symbols: Cognition and Commu-nication in Infancy, New York: Academic Press.

Bates, Elizabeth, Bretherton, Inge and Synder, Lynn (1988) From FirstWords to Grammar, Cambridge: Cambridge University Press.

Bauer, Laurie (1983) English Word-Formation, Cambridge: CambridgeUniversity Press.

—— (1988) Introducing Linguistic Morphology, Edinburgh: EdinburghUniversity Press.

—— (1998) Vocabulary, London: Routledge.Beekes, Robert S.P. (1995) Comparative Indo-European Linguistics,Amsterdam: Benjamins.

Bennett, Paul (1995) A Course in Generalized Phrase Structure Gram-mar, London: UCL Press.

Berko Gleason, Jean (1997) The Development of Language (4th edn),Boston, MA: Allyn & Bacon.

Berlin, Brent and Kay, Paul (1969) Basic Color Terms: Their Uni-versality and Evolution, Berkeley, CA: University of California Press.

Bex, Tony and Watts, Richard (eds) (1999) Standard English: TheWidening Debate, London: Routledge.

Bhatia, V.K. (1993) Analysing Genre: Language Use in ProfessionalSettings, London: Longman.

Biber, Douglas (2006) University Language: A Corpus-Based Study ofSpoken and Written Registers, Amsterdam: Benjamins.

Bickerton, Derek (1981) Roots of Language, Ann Arbor, MI: Karoma.—— (1984) ‘The Language Bioprogram Hypothesis’, Behavioral andBrain Sciences 7: 173–221.

—— (1990) Language and Species, Chicago, IL: University of ChicagoPress.

—— (1996) Language and Human Behaviour, London: UCL Press.Binnick, R. I. (1991) Time and the Verb: A Guide to Tense and Aspect,Oxford: Oxford University Press.

Blake, Barry (2001) Case (2nd edn), Cambridge: Cambridge UniversityPress.

Blake, N.F. (1996) A History of the English Language, London:Macmillan.

Blakemore, Diane (1992) Understanding Utterances, Oxford: Blackwell.

BIBLIOGRAPHY

331

Bloomfield, Leonard (1933) Language, New York: Holt, Rinehart &Winston.

Bloor, Thomas and Bloor, Meriel (2004) The Functional Analysis ofEnglish: A Hallidayan Approach (2nd edn), London: Arnold.

Blutner, Remhard and Zeevat, Henk (2003) Optimality Theory andPragmatics, Basingstoke: Palgrave Macmillan.

Bodmer, Frederick (1944) The Loom of Language, London: Allen &Unwin.

Bonvillain, Nancy (1993) Language, Culture, and Communication: TheMeaning of Messages, Englewood Cliffs, NJ: Prentice-Hall.

Borsley, Robert (1991) Syntactic Theory: A Unified Approach, London:Arnold.

—— (1996) Modern Phrase Structure Grammar, Oxford: Blackwell.—— (1999) Syntactic Theory: A Unified Approach (2nd edn), London:Arnold.

Bradford, Richard (1997) Stylistics, London: Routledge.Bratt-Paulston, C. and Tucker R. (2003) Sociolinguistics: The EssentialReadings. Oxford: Blackwell.

British Dyslexia Association (1996) Getting the Message Across: Dys-lexia: A Hundred Years of Progress? Birmingham: Questions.

Brown, Gillian and Yule, George (1983) Discourse Analysis, Cambridge:Cambridge University Press.

Brown, Keith and Miller, Jim (1991) Syntax: A Linguistic Introductionto Sentence Structure (2nd edn), London: HarperCollins.

Brown, Penelope, and Levinson, Stephen (1987) Politeness: Some Uni-versals in Language Usage, Cambridge: Cambridge University Press.

Brutt-Griffler, J. (2002) World English: A Study of its Development,Clevedon: Multilingual Matters.

Buckingham, Thomas and Yorkey, Richard (1984) Cloze Encounters:ESL Exercises in a Cultural Context, London: Prentice-Hall.

Burke, Peter (1993) The Art of Conversation, Cambridge: Polity Press.Burton-Roberts, Noel (1986) Analysing Sentences, London: Longman.Butler, Christopher S. (1985) Systemic Linguistics: Theory and Appli-cations, London: Batsford.

Butt, Miriam and King, Tracy (2000) Argument Realisation, Stanford,CA: CSLI.

Bybee, Joan L., Perkins, Revere and Pagliuca, William (1994) TheEvolution of Grammar: Tense, Aspect, and Modality in the Languagesof the World, Chicago, IL: University of Chicago Press.

Cameron, Deborah (1995) Verbal Hygiene, London: Routledge.Cameron, Deborah and Kulick, Don (2003) Language and Sexuality,Cambridge: Cambridge University Press.

Caplan, David (1992) Language: Structure, Processing and Disorders,Cambridge, MA: MIT Press.

Carr, Philip (1993) Phonology, London: Macmillan.Carter, Ronald (1997) Investigating English Discourse, London: Routledge.—— (2004) Language and Creativity: The Art of Common Talk,London: Routledge.

BIBLIOGRAPHY

332

Carter, Ronald, Goddard, Angela, Reah, Danuta, Sanger, Keith andBowring, Maggie (1997) Working with Texts: A Core Book for LanguageAnalysis, London: Routledge.

Catford, J.C. (1977) Fundamental Problems in Phonetics, Edinburgh:Edinburgh University Press.

—— (1988) A Practical Introduction to Phonetics, Oxford: OxfordUniversity Press.

Chafe, Wallace (1970) Meaning and the Structure of Language, Chicago,IL: University of Chicago Press.

—— (1994) Discourse, Consciousness and Time, Chicago, IL: Universityof Chicago Press.

Chambers, J.K. (2003) Sociolinguistic Theory: Linguistic Variation andits Social Significance (2nd edn), Oxford: Blackwell.

Chambers, J.K. and Trudgill, Peter (1998) Dialectology (2nd edn),Cambridge: Cambridge University Press.

Chambers, J.K., Trudgill, P. and Schilling-Estes, N. (eds) (2002) TheHandbook of Language Variation and Change, Oxford: Blackwell, pp.402–22.

Channell, Joanna (1994) Vague Language, Oxford: Oxford UniversityPress.

Chapman, Siobhan (2000) Philosophy for Linguists, London: Routledge.—— (2006) Thinking about Language, Basingstoke: Palgrave Macmillan.Chaudenson, R. (2001) Creolization of Language and Culture, London:Routledge.

Cheshire, Jenny and Stein, Dieter (eds) (1997) Taming the Vernacular:From Dialect to Written Standard Language, London: Longman.

Chilton, Paul (2003) Analysing Political Disocurse, London: Routledge.Chomsky, Noam (1977) The Logical Structure of Linguistic Theory,New York: Plenum Press.

—— (1995) The Minimalist Program, Cambridge, MA: MIT Press.—— (2000) The Architecture of Language, Oxford: Oxford UniversityPress.

Clark, John and Yallop, Colin (1995) An Introduction to Phonetics andPhonology (2nd edn), Oxford: Blackwell.

Coates, Jennifer (1983) The Semantics of the Modal Auxiliaries,London: Croom Helm.

—— (1996) Women Talk, Oxford: Blackwell.—— (ed.) (1998) Language and Gender: A Reader, Oxford: Blackwell.—— (2004) Women, Men and Language (3rd edn), London: Longman.Cobley, Paul (ed.) (2001) The Routledge Companion to Semiotics andLinguistics, London: Routledge.

Collins Cobuild (1990) English Grammar: London: HarperCollins.Collins, Beverley and Mees, Inger (2003) Practical Phonetics and Pho-nology: A Resource Book for Students, London: Routledge.

Comrie, Bernard (1976) Aspect, Cambridge: Cambridge University Press.—— (1985) Tense, Cambridge: Cambridge University Press.—— (1989) Language Universals and Linguistic Typology, Oxford:Blackwell.

BIBLIOGRAPHY

333

Comrie, Bernard, Matthews, Stephen and Polinsky, Maria (1997) TheAtlas of Languages, London: Bloomsbury.

Connell, R.W. (1995) Masculinities, Cambridge: Polity Press.Cook, Guy (1994) Discourse and Literature, Oxford: Oxford UniversityPress.

—— (2003) Applied Linguistics, Oxford: Oxford University Press.Cook, Vivian and Newson, Mark (1996) Chomsky’s Universal Grammar:An Introduction (2nd edn), Oxford: Blackwell.

Cooper, R. (1989) Language Planning and Social Change, Cambridge:Cambridge University Press.

Corbett, Greville (1991) Gender, Cambridge: Cambridge UniversityPress.

—— (2006) Agreement, Cambridge: Cambridge University Press.Corder, S. Pit (1975) Introducing Applied Linguistics, London: Penguin.Cotterill, Janet (ed.) (2002) Language in the Legal Process, Basingstoke:Palgrave Macmillan.

Coulmas, Florian (1996) The Blackwell Encyclopedia of Writing Systems,Oxford: Blackwell.

—— (2005) Sociolinguistics: The Study of Speakers’ Choices, Cam-bridge: Cambridge University Press.

Coulthard, Malcolm (1985) Introduction to Discourse Analysis (2ndedn), London: Longman.

Cowper, Elizabeth A. (1992) A Concise Introduction to SyntacticTheory, Chicago, IL: University of Chicago Press.

Croft, William (2002) Typology and Universals (2nd edn), Cambridge:Cambridge University Press.

Croft, William and Cruse, David Alan (2004) Cognitive Linguistics,Cambridge: Cambridge University Press.

Crowley, Terry (1997) An Introduction to Historical Linguistics (3rdedn), Oxford: Oxford University Press.

Crowley, Tony (1989) The Politics of Discourse: The Standard LanguageQuestion and British Cultural Debates, Basingstoke: Macmillan(published in the USA as The Politics of Standard English).

—— (1996) Language in History, London: Routledge.Cruse, David Alan (1986) Lexical Semantics, Cambridge: CambridgeUniversity Press.

Cruttenden, Alan (1986) Intonation, Cambridge: Cambridge UniversityPress.

Crystal, David (1975) The English Tone of Voice, London: Arnold.—— (1988) The English Language, London: Penguin.—— (1995) The Cambridge Encyclopedia of the English Language,Cambridge: Cambridge University Press.

—— (1996) Rediscover Grammar (2nd edn), London: Longman.—— (1997) The Cambridge Encyclopedia of Language, Cambridge:Cambridge University Press.

—— (2000) Language Death, Cambridge: Cambridge University Press.—— (2003) English as a Global Language (2nd edn), Cambridge:Cambridge University Press.

BIBLIOGRAPHY

334

Culicover, Peter (1997) Principles and Parameters, Oxford: OxfordUniversity Press.

Culler, Jonathan (1986) Ferdinand de Saussure (2nd edn), Ithaca, NY:Cornell University Press.

Cutting, Joan (2002) Pragmatics and Discourse: A Resource Book forStudents, London: Routledge.

Dahl, Osten (1985) Tense and Aspect Systems, Oxford: Blackwell.Dalby, Andrew (2003) Language in Danger, London: Penguin.Dale, Robert, Moisl, Hermann and Somers, Harold (eds) (2000)Handbook of Natural Language Processing, New York: MarcelDekker.

Daniels, Peter T. and Bright, William (eds) (1996) The World’s WritingSystems, Oxford: Oxford University Press.

Davies, Alan (1999) An Introduction to Applied Linguistics: FromPractice to Theory, Edinburgh: Edinburgh University Press.

Davis, Hayley G. (2001) Words: An Integrational Approach, Richmond:Curzon.

de Beaugrande, Robert (1994) ‘Text Linguistics’, in R.E. Asher andJ.M.Y. Simpson (eds) (1994) Encyclopedia of Language and Linguis-tics, Oxford: Pergamon, Vol. IX, pp. 4573–8.

de Beaugrande, Robert and Dressler, Wolfgang (1981) An Introductionto Text Linguistics, London: Longman.

de Fina, A. (2003) Identity in Narrative: A Study of Immigrant Discourse,Amsterdam: Benjamins.

Dekkers, Joost, van der Leeuw, Frank and van de Weijer, Jeroen (eds)(1999) Optimality Theory: Phonology, Syntax, and Acquisition,Oxford: Oxford University Press.

Denes, Peter B. and Pinson, Elliot N. (1993) The Speech Chain: ThePhysics and Biology of Spoken Language (2nd edn), New York: W.H.Freeman.

Deuchar, Margaret (1984) British Sign Language, London: Routledge.Dixon, R.M.W. (1994) Ergativity, Cambridge: Cambridge UniversityPress.

Duchan, Judith F. (2003) Frame Work in Language and Literacy: HowTheory Informs Practice, London: Guilford.

Duchan, Judith F., Bruder, Gail A. and Hewitt, L.E. (eds) (1995)Deixis in Narrative: A Cognitive Science Perspective, Hillsdale, NJ:Lawrence Erlbaum.

Dunkling, Leslie (1995) The Guinness Book of Names, Enfield: Guinness.Duranti, Alessandro (1997) Linguistic Anthropology, Cambridge: Cam-bridge University Press.

Eckert, Penelope and McConnell-Ginet, Sally (1992) ‘Think Practicallyand Look Locally: Language and Gender as Community-BasedPractice’, Annual Review of Anthropology 21: 461–90.

—— (1995) ‘Constructing Meaning, Constructing Selves: Snapshots ofLanguage, Gender and Class from Belten High’, in K. Hall and M.Bucholtz (eds) Gender Articulated: Language and the Socially Con-structed Self, London: Routledge, pp. 469–507.

BIBLIOGRAPHY

335

—— (1999) ‘New Generalizations and Explanations in Language andGender Research’, Language in Society 28 (2): 185–201.

—— (2003) Language and Gender, Cambridge: Cambridge UniversityPress.

Eckert, Penelope and Rickford, John R. (eds) (2001) Style and Socio-linguistic Variation, New York: Cambridge University Press.

Eco, Umberto (1976) A Theory of Semiotics, Bloomington, IN: IndianaUniversity Press.

Edwards, John (1994) Multilingualism, London: Penguin.Edwards, Viv (2004) Multilingualism in the English-Speaking World,Oxford: Blackwell.

Eelen, G. (2001) A Critique of Politeness Theories, Manchester: St Jerome.Eggins, Suzanne (2004) An Introduction to Systemic Functional Linguistics(2nd edn), London: Continuum.

Eggins, Suzanne and Slade, Diana (2004) Analysing Casual Conversation,London: Equinox.

Elgin, Suzette Haden (1983) What Is Linguistics? Englewood Cliffs, NJ:Prentice-Hall.

Emmitt, Marie, Komesaroff, Linda and Pollock, John (2006) Languageand Learning: An Introduction for Teaching (4th edn), Oxford: OxfordUniversity Press.

Epstein, David and Seely, T. Daniel (eds) (2002) Derivation and Expla-nation in the Minimalist Program, Oxford: Blackwell.

Fabb, Nigel (1997) Linguistics and Literature, Oxford: Blackwell.Fabbro, Franco and Asher, R.E. (eds) (1999) A Concise Encylopedia ofLanguage Pathology, Amsterdam: Elsevier.

Fairclough, Norman (ed.) (1992) Critical Language Awareness,London: Longman.

—— (1995) Critical Discourse Analysis: The Critical Study of Language,London: Longman.

—— (2001) Language and Power (2nd edn), London: Longman.—— (2003) Analysing Discourse: Textual Analysis for Social Research,London: Routledge.

Fauconnier, Gilles (1994) Mental Spaces (original in French as EspacesMentaux, 1984, Paris: Editions de Minuit), Cambridge: CambridgeUniversity Press.

—— (1997) Mappings in Thought and Language, Cambridge: Cam-bridge University Press.

Fauconnier, Gilles and Turner, Mark (2002) The Way We Think: Con-ceptual Blending and the Mind’s Hidden Complexities, New York:Basic Books.

Fennell, Barbara (2000) A History of English: A SociolinguisticApproach, Oxford: Blackwell.

Ferguson, Charles A. (1959) ‘Diglossia’, Word 15: 324–40. Reprinted inPier Paolo Giglioli (ed.) (1972) Language and Social Context, London:Penguin, pp. 232–51.

Fernando, C. (1996) Idiom and Idiomaticity, Oxford: Oxford UniversityPress.

BIBLIOGRAPHY

336

Field, John (2003) Psycholinguistics: A Resource Book for Students,London: Routledge.

Fillmore, Charles (1987) Fillmore’s Case Grammar, Heidelberg: Groos.Fishman, Joshua (ed.) (1999) Handbook of Language and Ethnic Identity,Oxford: Oxford University Press.

Fletcher, Paul and Macwhinney, Brian (eds) (1995) The Handbook ofChild Language, Oxford: Blackwell.

Fodor, Jerry A. (1983) Modularity of Mind: An Essay on Faculty Psy-chology, Cambridge, MA: MIT Press.

Foley, John (1993) The Guinness Encyclopedia of Signs and Symbols,Enfield: Guinness.

Foley, William A. (1997) Anthropological Linguistics: An Introduction,Oxford: Blackwell.

Fought, C. (2007) Language and Ethnicity, Cambridge: CambridgeUniversity Press.

Foulkes, Paul and Docherty, Gerry (eds) (1999) Urban Voices: Variationand Change in British Accents, London: Arnold.

Fowler, Roger (1996) Linguistic Criticism (2nd edn), Oxford: OxfordUniversity Press.

Fox, Anthony (1995) Linguistic Reconstruction, Oxford: Oxford UniversityPress.

Frawley, William (1992) Linguistic Semantics, Hillsdale, NJ: LawrenceErlbaum.

—— (ed.) (2006) The Expression of Modality, Berlin: Mouton deGruyter.

Fromkin, Victoria and Rodman, Robert (1998) An Introduction toLanguage (6th edn), Fort Worth, TX: Harcourt Brace.

Garman, Mike (1990) Psycholinguistics, Cambridge: Cambridge Uni-versity Press.

Garnham, Alan (1985) Psycholinguistics: Central Topics, London:Methuen.

Garside, R., Leech, G. and McEnery, A. (eds) (1997) Corpus Annota-tion: Linguistic Information from Computer Text Corpora, London:Longman.

Gavins, Joanna (2007) Text World Theory, Edinburgh: Edinburgh Uni-versity Press.

Gibbons, John (2003) Forensic Linguistics: An Introduction to Languagein the Justice System, Oxford: Blackwell.

Giegerich, Heinz J. (1992) English Phonology: An Introduction, Cam-bridge: Cambridge University Press.

Giora, Rachel (2003) On Our Mind: Salience, Context, and FigurativeLanguage, Oxford: Oxford University Press.

Givon, Talmy (1993) English Grammar: A Function-Based Introduction(2 vols), Amsterdam: Benjamins.

—— (1995) Functionalism and Grammar, Amsterdam: Benjamins.—— (2001) Syntax: An Introduction (rev. edn), Amsterdam: Benjamins.Gleason, H.A. (1961) An Introduction to Descriptive Linguistics (2ndedn), New York: Holt, Rinehart & Winston.

BIBLIOGRAPHY

337

Goatly, Andrew (1997) The Language of Metaphors, London: Routledge.Goldberg, Adele (2006) Constructions at Work: The Nature of Gen-eralization in Language, Oxford: Oxford University Press.

Goldstein, E. Bruce (2002) Sensation and Perception (6th edn), PacificGrove, CA: Wadsworth.

Goodluck, Helen (1991) Language Acquisition: A Linguistic Introduction,Oxford: Blackwell.

Gramley, Stephan and Patzold, Kurt-Michael (1992) A Survey ofModern English, London: Routledge.

Green, Jonathon (1996) Chasing the Sun: Dictionary-Makers and theDictionaries They Made, London: Jonathan Cape.

Green, Keith (ed.) (1995) New Essays in Deixis: Discourse, Narrative,Literature, Amsterdam: Rodopi.

Greenbaum, Sidney and Quirk, Randolph (1990) A Student’s Grammarof the English Language, London: Longman.

Greenberg, Joseph H. (1963) ‘Some Universals of Grammar with Parti-cular Reference to the Order of Meaningful Elements’, in J.H. Greenberg(ed.) Universals of Grammar, Cambridge, MA: MIT Press, pp. 73–113.

Grimes, Barbara F. (ed.) (2000) Ethnologue: Languages of the World(14th edn), Dallas, TX: Summer Institute of Linguistics.

Grinder, John T. and Elgin, Suzette Haden (1973) Guide to Transfor-mational Grammar: History, Theory, Practice, New York: Holt,Rinehart & Winston.

Grundy, Peter (1995) Doing Pragmatics, London: Arnold.Gumperz, John J. and Levinson, Stephen C. (eds) (1996) RethinkingLinguistic Relativity, Cambridge: Cambridge University Press.

Gussenhoven, Carlos and Jacobs, Haike (1998) Understanding Phonology,London: Arnold.

Gussmann, Edmund (2002) Phonology, Cambridge: Cambridge UniversityPress.

Haegeman, Liliane (1994) Introduction to Government and BindingTheory (2nd edn), Oxford: Blackwell.

Halliday, Michael A.K. (1989) Spoken and Written Language, Oxford:Oxford University Press.

—— (2004) An Introduction to Functional Grammar (3rd edn, revisedby Christian M.I.M. Matthiessen), London: Arnold.

Halliday, Michael A.K. and Hasan, Ruqaiya (1976) Cohesion in Eng-lish, London: Longman.

Harris, Randy Allen (1993) The Linguistics Wars, Oxford: OxfordUniversity Press.

Harris, Roxy (2006) New Ethnicities and Language Use, Basingstoke:Palgrave Macmillan.

Harris, Roy (1998) Introduction to Integrational Linguistics, Oxford:Elsevier.

Harris, Roy and Wolf, George (eds) (1998) Integrational Linguistics: AFirst Reader, Oxford: Elsevier.

Hawkins, John A. (1983) Word Order Universals, New York: AcademicPress.

BIBLIOGRAPHY

338

Hawkins, Peter (1984) Introducing Phonology, London: Hutchinson.Heggie, Lorie and Ordonez, Fernando (eds) (2004) Clitic and AffixCombinations: Theoretical Perspectives, Amsterdam: Benjamins.

Heil, John (1998) Philosophy of Mind: A Contemporary Introduction,London: Routledge.

Hendrick, Randall (ed.) (2003) Minimalist Syntax, Oxford: Blackwell.Hickmann, Maya and Robert, Stephane (eds) (2006) Space in Lan-guages: Linguistic Systems and Cognitive Categories, Amsterdam:Benjamins.

Hiraga, Masako (2005) Metaphor and Iconicity: A Cognitive Approachto Analysing Texts, Basingstoke: Palgrave Macmillan.

Hock, Hans Henrich (1986) Principles of Historical Linguistics, Berlin:Mouton de Gruyter.

Hock, Hans Henrich and Joseph, Brian D. (1996) Language History,Language Change and Language Relationship: An Introduction toHistorical and Comparative Linguistics, Berlin: Mouton de Gruyter.

Hockett, Charles F. (1958) A Course in Modern Linguistics, New York:Macmillan.

—— (1960) ‘The Origin of Speech’, Scientific American 203 (September):88–96.

Hoey, Michael (2005) Lexical Priming: A New Theory of Words andLanguage, London: Routledge.

Hofmann, Theo R. (1993) Realms of Meaning, London: Longman.Hofstadter, Douglas R. (1979) Godel, Escher, Bach: An Eternal GoldenBraid, Hassocks: Harvester.

—— (1985) ‘A Person Paper on Purity in Language’, in D.R. Hof-stadter, Metamagical Themas, New York: Basic Books, pp. 159–67.Reprinted in Deborah Cameron (ed.) (1990) The Feminist Critique ofLanguage: A Reader, London: Routledge, pp. 187–96.

Hogg, Richard and McCully, C.B. (1987) Metrical Phonology: ACoursebook, Cambridge: Cambridge University Press.

Holm, John (1988–9) Pidgins and Creoles (2 vols), Cambridge: Cam-bridge University Press.

—— (2000) An Introduction to Pidgins and Creoles, Cambridge: Cam-bridge University Press.

Holmes, Janet (1992) An Introduction to Sociolinguistics, London:Longman.

Holmes, Janet and Meyerhoff, Miriam (eds) (2003) The Handbook ofLanguage and Gender, Oxford: Blackwell.

Honey, John (1997) Language Is Power: The Story of Standard Englishand Its Enemies, London: Faber & Faber.

Horne, Kibbey M. (1966) Language Typology: 19th and 20th CenturyViews, Washington, DC: Georgetown University Press.

Hornstein, Norbert (1990) As Time Goes By: Tense and UniversalGrammar, Cambridge, MA: MIT Press.

Hornstein, Norbert, Nunes, Jairo and Grohmann, Kleanthes (2005)Understanding Minimalism, Cambridge: Cambridge University Press.

Horrocks, Geoffrey (1987) Generative Grammar, London: Longman.

BIBLIOGRAPHY

339

Huddleston, Rodney (1984) Introduction to the Grammar of English,Cambridge: Cambridge University Press.

Hudson, Richard A. (1984) Invitation to Linguistics, Oxford: Blackwell.—— (1995) Word Meaning, London: Routledge.—— (1996) Sociolinguistics (2nd edn), Cambridge: Cambridge UniversityPress.

—— (1998) English Grammar, London: Routledge.Hughes, Arthur and Trudgill, Peter (1996) English Accents and Dialects:An Introduction to Social and Regional Varieties of English in theBritish Isles (3rd edn), London: Arnold.

Hurford, James R. (1994) Grammar: A Student’s Guide, Cambridge:Cambridge University Press.

Hurford, James R. and Heasley, Brendan (1983) Semantics: A Course-book, Cambridge: Cambridge University Press.

Hyman, Larry M. (1975) Phonology: Theory and Analysis, New York:Holt, Rinehart & Winston.

Ilson, Robert (ed.) (1986) Lexicography: An Emerging InternationalProfession, Manchester: Manchester University Press.

Ingram, David (1989) First Language Acquisition: Method, Descriptionand Explanation, Cambridge: Cambridge University Press.

Ingram, John (2007) Neurolinguistics, Cambridge: Cambridge UniversityPress.

Jackendoff, Ray (1993) Patterns in the Mind: Language and HumanNature, New York: Harvester Wheatsheaf.

Jackson, Howard (2000) Words, Meaning and Vocabulary: An Introductionto Modern English Lexicology, London: Cassell.

—— (2002) Grammar and Vocabulary: A Resource Book for Students,London: Routledge.

Jeffries, Lesley (2006) Discovering Language: The Structure of ModernEnglish, Basingstoke: Palgrave Macmillan.

Jenkins, J. (2003) World Englishes: A Resource Book for Students,London: Routledge.

Johnson, K. (1997) Acoustic and Auditory Phonetics, Oxford: Blackwell.Johnson-Laird, Philip H. (1983) Mental Models: Towards a CognitiveScience of Language, Inference and Consciousness, Cambridge: Cam-bridge University Press.

—— (1993) The Computer and the Mind: An Introduction to CognitiveScience (2nd edn), London: Fontana.

Johnstone, Barbara (2000) Qualitative Methods in Sociolinguistics, NewYork: Oxford University Press.

Joseph, J. (2004) Language and Identity, Basingstoke: Palgrave.Joshi, R. Malatesha and Aaron, P.G. (eds) (2006) Handbook of Ortho-graphy and Literacy, Mahwah, NJ: Lawrence Erlbaum.

Jurafsky, Dan and Martin, James (2000) Speech and Language Proces-sing: An Introduction to Natural Language Processing, ComputationalLinguistics, and Speech Recognition, London: Prentice-Hall.

Kager, Reni (1999) Optimality Theory, Cambridge: Cambridge UniversityPress.

BIBLIOGRAPHY

340

Kaplan, Ronald and Bresnan, Joan (1982) ‘Lexical-Functional Gram-mar: A Formal System for Grammatical Representation’, in J. Bresnan(ed.), The Mental Representation of Grammatical Relations, Cambridge,MA: MIT Press, pp. 173–281.

Katamba, Francis (1989) An Introduction to Phonology, London:Longman.

—— (1994) English Words, London: Routledge.Katamba, Francis and Stonham, John T. (2006) Morphology, Basingstoke:Palgrave Macmillan.

Kempson, Ruth M. (1977) Semantic Theory, Cambridge: CambridgeUniversity Press.

Keyser, Samuel Jay and Postal, Paul M. (1976) Beginning EnglishGrammar, New York: Harper & Row.

Kilby, David (1984) Descriptive Syntax and the English Verb, London:Croom Helm.

Klaiman, M.H. (1991) Grammatical Voice, Cambridge: CambridgeUniversity Press.

Klima, Edward S. and Bellugi, Ursula, (1979) The Signs of Language,Cambridge, MA: Harvard University Press.

Knowles, Gerald (1987) Patterns of Spoken English, London: Longman.Kouwenberg, S. and Singler, J.V. (eds) (2005) The Handbook of Pidginand Creole Studies, Malden, MA: Blackwell.

Kovecses, Zoltan (2005) Metaphor in Culture: Universality and Variation,Cambridge: Cambridge University Press.

Krauss, Michael (1992) ‘The World’s Languages in Crisis’, Language68: 4–10.

Kreidler, Charles W. (1989) The Pronunciation of English, Oxford:Blackwell.

—— (1998) Introducing English Semantics, London: Routledge.Kroeger, Paul (2004) Analysing Syntax, Cambridge: Cambridge UniversityPress.

—— (2005) Analyzing Grammar, Cambridge: Cambridge UniversityPress.

Kuhn, Jonas (2003) Optimality-Theoretic Syntax: A DeclarativeApproach, Stanford, CA: CSLI.

Labov, William (1972) Sociolinguistic Patterns, Philadelphia, PA: Uni-versity of Pennsylvania Press.

—— (1975) What Is a Linguistic Fact? Lisse: Peter de Ridder Press.—— (1994) Principles of Linguistic Change, Vol. I: Internal Factors,Oxford: Blackwell.

Ladefoged, Peter (1971) Preliminaries to Linguistic Phonetics, Chicago,IL: University of Chicago Press.

—— (1993) A Course in Phonetics (3rd edn), Fort Worth, TX: HarcourtBrace Jovanovich.

Ladefoged, Peter and Maddieson, Ian (1996) The Sounds of the World’sLanguages, Oxford: Blackwell.

Lakoff, George (1987) Women, Fire, and Dangerous Things, Chicago,IL: University of Chicago Press.

BIBLIOGRAPHY

341

Lakoff, George and Johnson, Mark (1980) Metaphors We Live By,Chicago, IL: University of Chicago Press.

Landau, Sidney I. (1984) Dictionaries: The Art and Craft of Lexicography,New York: Charles Scribner’s Sons.

Langacker, Ronald W. (1987–91) Foundations of Cognitive Grammar,Berlin: Mouton de Gruyter.

—— (1990) Concept, Image and Symbol: The Cognitive Basis of Grammar,Berlin: Mouton de Gruyter.

Langlotz, Andreas (2006) Idiomatic Creativity: A Cognitive-LinguisticModel of Idiom Representation and Idiom Variation in English,Amsterdam: Benjamins.

Large, Andrew (1985) The Artificial Language Movement, Oxford:Blackwell.

Lasnik, Howard (2003) Minimalist Investigations in Linguistic Theory,London: Routledge.

Lasnik, Howard, Uriagereka, Juan and Boeckx, Cedric (2005) A Coursein Minimalist Syntax: Foundations and Prospects, Oxford: Blackwell.

Lass, Roger (1984) Phonology, Cambridge: Cambridge University Press.Laver, John (1994) Principles of Phonetics, Cambridge: CambridgeUniversity Press.

Leech, Geoffrey (1974) Semantics, London: Penguin.Leeds-Hurwitz, Wendy (1993) Semiotics and Communication: Signs,Codes, Culture, Hillsdale, NJ: Lawrence Erlbaum.

Lehiste, Ilse (1970) Suprasegmentals, Cambridge, MA: MIT Press.Lehmann, Winfred P. (1992) Historical Linguistics (3rd edn), London:Routledge.

—— (1993) Theoretical Bases of Indo-European Linguistics, London:Routledge.

—— (ed.) (1967) A Reader in Nineteenth-Century Historical Indo-European Linguistics, Bloomington, IN: University of Indiana Press.

Leith, Dick (1997) A Social History of English (2nd edn), London:Routledge.

Lepschy, Giulio (1970) A Survey of Structural Linguistics, London:Faber & Faber.

—— (ed.) (1994) History of Linguistics, Vol. II: Classical and MedievalLinguistics, London: Longman.

Levinson, Stephen (1983) Pragmatics, Cambridge: Cambridge UniversityPress.

Li, Wei (ed.) (2000) The Bilingualism Reader, London: Routledge.Llamas, Carmen, Mullany, Louise and Stockwell, Peter (eds) (2007)The Routledge Companion to Sociolinguistics, London: Routledge.

Lockwood, W.B. (1969) Indo-European Philology, London: Hutchinson.—— (1972) A Panorama of Indo-European Languages, London:Hutchinson.

Lodge, R. Anthony (1993) French: From Dialect to Standard, London:Routledge.

Lucy, John (1968) Introduction to Theoretical Linguistics, Cambridge:Cambridge University Press.

BIBLIOGRAPHY

342

—— (1991) Chomsky (3rd edn), London: Fontana.—— (1992) Grammatical Categories and Cognition: A Case Study of theLinguistic Relativity Hypothesis, Cambridge: Cambridge UniversityPress.

—— (1995) Linguistic Semantics: An Introduction, Cambridge: CambridgeUniversity Press.

Lucy, John and Shweder, Richard (1979) ‘Whorf and His Critics: Lin-guistic and Nonlinguistic Influences on Colour Memory’, AmericanAnthropologist 81: 581–615.

Lust, Barbara (2006) Child Language, Cambridge: Cambridge UniversityPress.

Heller, H. (ed.) (2007) Bilingualism, London: Palgrave.McArthur, Tom (1986) Worlds of Reference: Lexicography, Learningand Language from the Clay Tablet to the Computer, Cambridge:Cambridge University Press.

McCarthy, John (2001) A Thematic Guide to Optimality Theory, Cam-bridge: Cambridge University Press.

—— (ed.) (2004) Optimality Theory in Phonology: A Reader, Oxford:Blackwell.

McCarthy, Michael (1991) Discourse Analysis for Language Teachers,Cambridge: Cambridge University Press.

McCarthy, Michael and Carter, Ronald (1993) Language as Discourse:Perspectives for Language Teaching, London: Longman.

McCrum, Robert, Cran, William and MacNeil, Robert (1992) TheStory of English (2nd edn), London: Faber & Faber/BBC Books.

McEnery, Tony and Wilson, Andrew, (2001) Corpus Linguistics: AnIntroduction (2nd edn), Edinburgh: Edinburgh University Press.

McEnery, Tony, Xiao, Richard and Tono, Yukio (2006) Corpus-BasedLanguage Studies: An Advanced Resource Book, London: Routledge.

MacKenzie, Ian (2002) Paradigms of Reading: Relevance Theory andDeconstruction, Houndmills: Palgrave Macmillan.

McMahon, April (1994) Understanding Language Change, Cambridge:Cambridge University Press.

—— (2000) Change, Chance, and Optimality, Oxford: Oxford UniversityPress.

McMenamin, Gerald R. (ed.) (2002) Forensic Linguistics: Advances inForensic Stylistics, Boca Raton, FL: CRC Press.

Macwhinney, Brian and Bates, Elizabeth (eds) (1989) The Cross-linguistic Study of Linguistic Processing, Cambridge: CambridgeUniversity Press.

Mahlberg, Michaela (2005) English General Nouns: A Corpus Theore-tical Approach, Amsterdam: Benjamins.

Mallory, J.P. (1989) In Search of the Indo-Europeans, London: Thames& Hudson.

Malmkjær, Kirsten (ed.) (2004) The Linguistics Encyclopedia (2nd edn),London: Routledge.

Martin, Robert M. (1987) The Meaning of Language, Cambridge, MA:MIT Press.

BIBLIOGRAPHY

343

Masterman, Margaret and Wilks, Yorick (eds) (2005) Language, Cohe-sion and Form, Cambridge: Cambridge University Press.

Matthews, Peter (1979) Generative Grammar and Linguistic Compe-tence, London: Allen & Unwin.

—— (1981) Syntax, Cambridge: Cambridge University Press.—— (1991) Morphology (2nd edn), Cambridge: Cambridge UniversityPress.

Maybin, Janet and Mercer, Neil (1996) Using English: From Conversationto Canon, London: Routledge/Open University.

Maybin, Janet and Swann, Joan (eds) (2006) The Art of English:Everyday Creativity, Basingstoke: Palgrave Macmillan.

Mey, Jacob (1993) Pragmatics, Oxford: Blackwell.Miles, Dorothy (1988) British Sign Language: A Beginner’s Guide,London: BBC Books.

Mills, Sara (2003) Gender and Politeness, Cambridge: Cambridge Uni-versity Press.

—— (2004) Discourse, New York: Routledge.Milroy, James (1992) Linguistic Variation and Change, Oxford: Blackwell.Milroy, James and Milroy, Lesley (eds) (1993) Real English: TheGrammar of English Dialects in the British Isles, London: Longman.

—— (1997) Authority in Language: Investigating Standard English, (3rdedn), London: Routledge.

Milroy, Lesley (1987a) Language and Social Networks (2nd edn),Oxford: Blackwell.

—— (1987b) Observing and Analysing Natural Language: A CriticalAccount of Sociolinguistic Method, Oxford: Blackwell.

Milroy, Lesley and Gordon, Matthew (2003) Sociolinguistics. Methodand Interpretation, Oxford: Blackwell.

Morgan, Gary and Woll, Bencie (2002) Directions in Sign LanguageAcquisition, Philadelphia, PA: Benjamins.

Nettle, D. and Romaine, Suzanne (2000) Vanishing Voices: The Extinctionof the World’s Languages. Oxford: Oxford University Press.

Nevalainen, Terttu, Klemola, Juhani and Laitinen, Mikko (eds) (2006)Types of Variation: Diachronic, Dialectical and Typological Interfaces,Amsterdam: Benjamins.

Newmeyer, Frederick J. (1983) Grammatical Theory: Its Limits andPossibilities, Chicago, IL: University of Chicago Press.

Nida, Eugene (1975) Componential Analysis of Meaning, The Hague:Mouton.

Nivre, Joakim (2006) Inductive Dependency Parsing, Dordrecht: Springer.Nofsinger, Robert E. (1991) Everyday Conversation, Newbury Park, CA:Sage.

O’Grady, William, Dobrovolsky, Michael and Katamba, Francis (1996)Contemporary Linguistics: An Introduction, London: Longman.

Odden, David (2005) Introducing Phonology, Cambridge: CambridgeUniversity Press.

Oller, John W. and Jonz, Jon (eds) (1994) Cloze and Coherence,London: Associated University Presses.

BIBLIOGRAPHY

344

Ong, Walter (1982) Orality and Literacy: The Technologising of theWord, London: Methuen.

Ouhalla, Jamal (1994) Introducing Transformational Grammar, London:Arnold.

Owens, Robert E., Jr. (1996) Language Development: An Introduction(4th edn), Boston, MA: Allyn & Bacon.

Padden, Carol A. (1988) ‘Grammatical Theory and Signed Languages’,in Frederick J. Newmeyer (ed.) Linguistics: The Cambridge Survey,Vol. II: Linguistic Theory: Extensions and Implications, Cambridge:Cambridge University Press, pp. 250–66.

Palmer, Frank (1971) Grammar, London: Penguin.—— (1974) The English Verb, London: Longman.—— (1976) Semantics: A New Outline, Cambridge: Cambridge UniversityPress.

—— (1979) Modality and the English Modals, London: Longman.—— (1994) Grammatical Roles and Relations, Cambridge: CambridgeUniversity Press.

—— (2001) Mood and Modality (2nd edn), Cambridge: CambridgeUniversity Press.

Palmer, Gary B. (1996) Toward a Theory of Cultural Linguistics, Austin,TX: University of Texas Press.

Partridge, Eric (1970) Slang To-Day and Yesterday, London: Routledge.—— (2002) A Dictionary of Slang and Unconventional English (6th edn,ed. by Paul Beale), London: Routledge.

Peirce, Charles Sanders (1991) Peirce on Signs: Writings on Semiotics(ed. by James Hoopes), Chapel Hill, NC: University of North CarolinaPress.

Pennycook, Alastair (2001) Critical Applied Linguistics: A CriticalIntroduction, Mahwah, NJ: Lawrence Erlbaum.

Piattelli-Palmarini, Massimo (ed.) (1979) Language and Learning: TheDebate Between Jean Piaget and Noam Chomsky, London: Routledge.

Pike, Kenneth Lee (1967) Language in Relation to a Unified Theory ofStructure of Human Behavior (2nd edn), The Hague: Mouton.

—— (1982) Linguistic Concepts: An Introduction to Tagmemics. Lincoln,NE: University of Nebraska Press.

Pinker, Steven (1994) The Language Instinct: The New Science of Lan-guage and Mind, London: Allen Lane/Penguin.

Polanyi, Livia (1985) Telling the American Story: A Structural andCultural Analysis of Conversation, Norwood, MA: Ablex.

Pollard, Carl and Sag, Ivan A. (1994) Head-Driven Phrase StructureGrammar, Chicago, IL: University of Chicago Press.

Poole, Geoffrey (2002) Syntactic Theory, Basingstoke: Palgrave.Poyatos, Fernando (1993) Paralanguage: A Linguistic and Inter-disciplinary Approach to Interactive Speech and Sound, Amsterdam:Benjamins.

—— (2002) Paralanguage, Kinesics, Silence, Personal and EnvironmentalInteraction, Amsterdam: Benjamins.

Pridham, F (2001) The Language of Conversation, London: Routledge.

BIBLIOGRAPHY

345

Propp, Vladimir (1968) Morphology of the Folktale, Austin, TX: Uni-versity of Texas Press.

Pullum, Geoffrey K. (1991) The Great Eskimo Vocabulary Hoax andOther Irreverent Essays on the Study of Language, Chicago, IL: Uni-versity of Chicago Press.

Pullum, Geoffrey K. and Ladusaw, William A. (1996) Phonetic SymbolGuide (2nd edn), Chicago, IL: University of Chicago Press.

Radford, Andrew (1988) Transformational Grammar: A First Course,Cambridge: Cambridge University Press.

—— (1997) Syntactic Theory and the Structure of English: A MinimalistApproach, Cambridge: Cambridge University Press.

—— (2004)Minimalist Syntax, Cambridge: Cambridge University Press.Rampton, Ben (1995) Crossing: Language and Ethnicity Among Ado-lescents, London: Longman.

Rauh, G. (ed.) (1983) Essays on Deixis, Tubingen: Gunter Narr Verlag.Renkema, J. (2004) Introduction to Discourse Studies, Philadelphia, PA:Benjamins.

Richards, Jack C., Platt, John and Platt, Heidi (1992) Dictionary ofLanguage Teaching and Applied Linguistics (2nd edn), London:Longman.

Roach, Peter (1991) English Phonetics and Phonology: A PracticalCourse (2nd edn), Cambridge: Cambridge University Press.

Robins, R.H. (1997) A Short History of Linguistics (4th edn), London:Longman.

Robinson, Andrew (1995) The Story of Writing, London: Thames &Hudson.

Robson, Mark and Stockwell, Peter (2005) Language in Theory: AResource Book for Students, London: Routledge.

Romaine, Suzanne (1988) Pidgin and Creole Languages, London:Longman.

—— (1994) Language in Society: An Introduction to Sociolinguistics,Oxford: Oxford University Press.

—— (1995) Bilingualism (2nd edn), Oxford: Blackwell.Ronen, Ruth (1994) Possible Worlds in Literary Theory, Cambridge:Cambridge University Press.

Ryan, Marie-Laure (1991) Possible Worlds, Artificial Intelligence andNarrative Theory, Bloomington, IN: Indiana University Press.

Sadock, Jerrold (1991) Autolexical Syntax: A Theory of Parallel Gram-matical Representations, Chicago, IL: University of Chicago Press.

Saeed, John I. (1997) Semantics, Oxford: Blackwell.Sampson, Geoffrey (1975) The Form of Language, London: Weidenfeld& Nicolson.

—— (1980) Schools of Linguistics, London: Hutchinson.—— (1985) Writing Systems, London: Hutchinson.—— (1997) Educating Eve: The ‘Language Instinct’ Debate, London:Cassell.

Sampson, Geoffrey and McCarthy, Diana (eds) (2005) Corpus Linguistics:Readings in a Widening Discipline, London: Continuum.

BIBLIOGRAPHY

346

Saville-Troike, M. (2003) The Ethnography of Communication (3rd edn),Oxford: Blackwell.

Schank, Roger C. (1982) Dynamic Memory: A Theory of Reminding andLearning in Computers and People, Cambridge: Cambridge UniversityPress.

—— (1986) Explanation Patterns, Hillsdale, NJ: Lawrence Erlbaum.Schank, Roger C. and Abelson, Robert (1977) Scripts, Plans, Goals andUnderstanding, Hillsdale, NJ: Lawrence Erlbaum.

Schiffrin, Deborah (1994) Approaches to Discourse, Oxford: Blackwell.Schmitt, Norbert (ed.) (2002) An Introduction to Applied Linguistics,London: Arnold.

Schmitt, Norbert and McCarthy, Michael (1997) Vocabulary: Description,Acquisition and Pedagogy, Cambridge: Cambridge University Press.

Sebba, Mark (1997) Contact Languages: Pidgins and Creoles, London:Macmillan.

Sebeok, Thomas (1984) Encyclopedic Dictionary of Semiotics (3 vols),Berlin: Mouton de Gruyter.

—— (1994) An Introduction to Semiotics, London: Pinter.Sells, Peter (1985) Lectures on Contemporary Syntactic Theories, Stanford,CA: CSLI.

Semino, Elena (1997) Language and World Creation in Poems and OtherTexts, London: Longman.

Shockey, Linda (2003) Sound Patterns of Spoken English, Oxford:Blackwell.

Shuy, Roger W. (2005) Creating Language Crimes: How Law Enforce-ment Uses (and Misuses) Language, Oxford: Oxford UniversityPress.

—— (2006) Linguistics in the Courtroom: A Practical Guide, Oxford:Oxford University Press.

Siewierska, Anna (1991) Functional Grammar, London: Routledge.Silverman, David (1993) Interpreting Qualitative Data, London: Sage.Simpson, Paul (1993) Language, Ideology and Point of View, London:Routledge.

—— (1996) Language through Literature: An Introduction, London:Routledge.

—— (2004) Stylistics: A Resource Book for Students, London: Routledge.Sinclair, John (1991) Corpus, Concordance and Collocation, Oxford:Oxford University Press.

Sinclair, John and Carter, Ronald (eds) (2004) Trust the Text: Language,Corpus and Discourse, London: Routledge.

Sinclair, John and Coulthard, Malcolm (1975) Towards an Analysis ofDiscourse: The English Used by Teachers and Pupils, Oxford: OxfordUniversity Press.

Sloane, Thomas O. (ed.) (2001) Encyclopedia of Rhetoric, Oxford: OxfordUniversity Press.

Smith, Cath (1990) Signs Make Sense: A Guide to British Sign Language,London: Souvenir Press.

Sommerstein, Alan H. (1977) Modern Phonology, London: Arnold.

BIBLIOGRAPHY

347

Sperber, Dan and Wilson, Deirdre (1995) Relevance: Communicationand Cognition (2nd edn), Oxford: Blackwell Publishers.

Stainton, Robert J. (1996) Philosophical Perspectives on Language,Peterborough, OR: Broadview.

Steinberg, Danny D. (1993) An Introduction to Psycholinguistics,London: Longman.

Stillings, Neil A. (1987) Cognitive Science: An Introduction, Cambridge,MA: MIT Press.

Stockwell, Peter (2002a) Sociolinguistics: A Resource Book for Students,London: Routledge.

—— (2002b) Cognitive Poetics: An Introduction, London: Routledge.—— (2003) ‘Schema Poetics and Speculative Cosmology’, Languageand Literature 12(3): 252–71.

—— (2007) Texture: A Cognitive Aesthetics of Reading, Edinburgh:Edinburgh University Press.

Sturrock, John (1993) Structuralism (2nd edn), London: Fontana.Szemerenyi, Oswald J.L. (1996) Introduction to Indo-European Linguistics,Oxford: Clarendon Press.

Tannen, Deborah (1991) You Just Don’t Understand: Women and Menin Conversation, London: Virago.

Tarleton, R. (1988) Learning and Talking: A Practical Guide to OracyAcross the Curriculum, London: Routledge.

Taylor, John R. (2003) Linguistic Categorization, Oxford: Oxford Uni-versity Press.

Tench, Paul (1996) The Intonation Systems of English, London: Cassell.Thibault, Paul J. (1994) ‘Intertextuality’, in R.E. Asher and J.M.Y.Simpson (eds), Encyclopedia of Language and Linguistics, Oxford:Pergamon, Vol. IV, pp. 1751–4.

Thomas, Jenny (1995) Meaning in Interaction, London: Longman.Thompson, Geoff (2004) Introducing Functional Grammar (2nd edn),London: Arnold.

Thornborrow, Joanna and Wareing, Shan (1998) Patterns in Language:An Introduction to Language and Literary Style, London: Routledge.

Tobin, Yishai (1990) Semiotics and Linguistics, London: Longman.Tognini-Bonelli, Elena (2001) Corpus Linguistics at Work, Amsterdam:Benjamins.

Toolan, Michael (1988) Narrative: A Critical Linguistic Introduction,London: Routledge.

—— (1994a) ‘Narrative: Linguistic and Structural Theories’, in R.E.Asher and J.M.Y. Simpson (eds), Encyclopedia of Language and Lin-guistics, Oxford: Pergamon, vol. V, pp. 2679–96.

—— (1994b) ‘Narrative, Natural’, in R.E. Asher and J.M.Y. Simpson(eds), Encyclopedia of Language and Linguistics, Oxford: Pergamon,Vol. V, pp. 2696–2701.

—— (1996) Total Speech: An Integrational Linguistic Approach toLanguage, Durham, NC: Duke University Press.

Trask, R.L. (1994) Language Change, London: Routledge.—— (1995) Language: The Basics, London: Routledge.

BIBLIOGRAPHY

348

—— (1996) Historical Linguistics, London: Arnold.Trudgill, Peter (1995) Sociolinguistics (2nd edn), London: Penguin.Trudgill, Peter and Hannah, Jean (1994) International English: A Guideto the Varieties of Standard English (3rd edn), London: Arnold.

Turing, Alan (1950) ‘Computing Machinery and Intelligence’, Mind 59(236, October): 433–60.

Turner, G.W. (1973) Stylistics, London: Penguin.Ungerer, F. and Schmid, H.-J. (1996) An Introduction to Cognitive Lin-guistics, London: Longman.

Valli, Clayton and Lucas, Ceil (2001) The Linguistics of American SignLanguage: An Introduction (3rd edn), Washington, DC: GallaudetUniversity Press.

van Peer, Willie (1986) Stylistics and Psychology: Investigations ofForegrounding, London: Croom Helm.

—— (1994) ‘Text’, in R.E. Asher and J.M.Y. Simpson (eds), Encyclopediaof Language and Linguistics, Oxford: Pergamon, Vol. IX, pp. 4564–8.

van Valin, Robert (2001) An Introduction to Syntax, Cambridge: Cam-bridge University Press.

van Valin, Robert D. and LaPolla, Randy (1997) Syntax: Structure,Meaning and Function, Cambridge: Cambridge University Press.

Verschueren, Jef (1999) Understanding Pragmatics, London: Arnold.Wales, Katie (2001) A Dictionary of Stylistics (2nd edn), London:Pearson.

Wallman, Joel (1992) Aping Language, Cambridge: Cambridge UniversityPress.

Wardhaugh, Ronald (2005) An Introduction to Sociolinguistics (5thedn), Oxford: Blackwell.

—— (1987) Languages in Competition, Oxford: Blackwell.Watts, Richard, Ide, Sachiko and Ehlich, Konrad (eds) (2005) Politeness inLanguage: Studies in its History, Theory, and Practice (2nd edn), NewYork: Mouton de Gruyter.

Weinreich, Uriel, Labov, William and Herzog, Marvin I. (1968)‘Empirical Foundations for a Theory of Language Change’, in W.P.Lehmann and Yakov Malkiel (eds) Directions for Historical Linguistics,Austin, TX: University of Texas Press, pp. 95–188.

Wells, John (1982) Accents of English (3 vols), Cambridge: CambridgeUniversity Press.

Wenger, E. (1998) Communities of Practice, Cambridge: CambridgeUniversity Press.

Wengraf, T. (2001) Qualitative Social Interviewing: Biographic Narrativeand Semi-Structured Methods. London: Sage.

Werth, Paul (1999) Text Worlds: Representing Conceptual Space inDiscourse, London: Longman.

Whaley, Lindsay J. (1997) Introduction to Typology: The Unity andDiversity of Language, London: Sage.

Whorf, Benjamin Lee (1956) Language, Thought and Reality: SelectedWritings of Benjamin Lee Whorf, ed. by J.B. Carroll, Cambridge,MA: MIT Press.

BIBLIOGRAPHY

349

Wierzbicka, Anna (1996) Semantics: Primes and Universals, Oxford:Oxford University Press.

Wodak, Ruth (1996) Disorders of Discourse, London: Longman.Wodak, Ruth and Chilton, Paul (2005) Methods of Critical DiscourseAnalysis. London: Sage.

Wolf, Florian and Gibson, Edward (2006) Coherence in Natural Language:Data Structures and Applications, Cambridge, MA: MIT Press.

Wooffitt, R. (2005) Conversation Analysis and Discourse Analysis: AComparative and Critical Introduction, London: Sage.

Wright, S. (2004) Language Policy and Language Planning: FromNationalism to Globalisation, Basingstoke: Palgrave.

Yip, Moira (2002) Tone, Cambridge: Cambridge University Press.Yule, George (1996) Pragmatics, Oxford: Oxford University Press.Zentella, A.C. (1997) Growing up Bilingual, Oxford: Blackwell.

BIBLIOGRAPHY

350

Index

Note: page numbers in bold indicate the entry for a subject.

abbreviation 201abessive case 36ablative case 36absolute transitive 306abstract sound unit 213accent 3–4, 9–10, 40, 72–3, 81,83, 93, 115, 125, 140, 164,170, 202, 213–14, 231–2, 250,263, 272

acceptability 83accommodation 164accusative case 11, 36, 103, 35–6,75, 203, 206

acoustic features 79acoustic phonetics 215acquired dyslexia 82acrolect 60acronym 326active vocabulary 151active voice 319–20adjacency pair 4–5, 309adjective 5–6, 8, 11, 96, 106–7,110, 118, 175, 177, 183, 187–8,203, 205–8, 218, 235, 278, 284,288, 308

adjective-complement clause 283adjunct 24, 329adposition 229adverb 6–8, 38, 58, 110, 177–8,207–8, 230, 233, 283

adverb of manner 8adverb of place 8adverb of time 8adverbial clause 38, 50, 66, 283adverbial phrase 24, 59, 66, 110

affix 6, 8–9, 35, 68–9, 118, 134,151, 164, 177–8, 224, 325–6

affricate 9–10, 14, 51, 124, 162–3African-American (Vernacular)

English 26, 33agent 220, 251–2, 282, 306, 316, 320agglutinating language 310agraphia 82agreement 10–11, 29, 39, 67–8, 90,

102–3, 108, 190, 317airstream mechanism 12–13,

213, 319Aktionsart 27alexia 82aliteracy 159allative case 36allomorph 177allophone 214alphabet 108–9, 123, 150, 166,

201–2, 238, 271, 327–8alternant 13alternation 13–14alveolar 115, 222Alzheimer’s disease 139–40ambiguity 14–15, 64, 257American Sign Language 260American structuralists 31, 79, 157,

217, 248, 252, 277, 283–4American transcription 124analogy 15–16, 202anaphor 16–17, 41, 55, 96, 234anaphora 16–17ancestral language 141animal communication 17–18, 23,

70, 81, 141, 201, 236

351

anomalous 264, 323anomia 139antecedent 16anthropological linguistics 18–19,

40, 43, 85, 129, 155, 181anthroponym 197antonymy 256Apalai 266aphasia 19–21, 82, 133–4, 139–40,161, 185, 270

apical 222Appalachian English 145apparent time 21applied linguistics 21–2, 39, 76, 82,93, 127, 270

appropriateness 43approximant 52, 162–3Apurina 30Arabic 23, 74, 88, 109, 122, 152,191, 201, 247, 273, 328

arbitrariness 22–3, 70, 111–12, 119,155, 165, 169, 266

arbitrary control 54arcuate fasciculus 20, 34areal linguistics 153argument 23–4, 48, 75, 115–16,228, 282, 317

Aristotle 228, 247, 289articulatory features 79articulatory phonetics 215artificial language 24–5, 168, 184artlang 25ascriptive sentence 59ASL see American Sign Languageaspect 26–7, 29, 107, 294–5Aspects model 304association 51atelic 27atomism 277attributive expression 246auditory phonetics 215Austin, J.L. 210, 212, 267Australian languages 29, 37, 53, 63,88, 97, 159, 173, 182, 192, 222,229, 277, 310

autolexical grammar 27autonomous phoneme 215autonomy 27–8, 99, 141autosegmental phonology 217auxiliary 29, 39, 59, 90, 279, 317

auxlang 25aversive case 36

back-channel 87back-formation 191Bakhtin, Mikhail 101, 125Bamileke-Dschang 294Basic English 25basic word order 30, 32, 310–11basilect 60Basque 10–11, 22–3, 26, 30, 35, 37,

73, 75, 103, 112–14, 117, 128,144–5, 171–3, 196, 229, 239, 266,273, 283

Bates, Elizabeth 99Baudouin de Courtenay, Jan 156,

177, 214, 216bee-dance 78behaviourism 30–31, 131, 167Bell, Alexander Melville 216Bella Coola 218Bellugi, Ursula 261Berlin, Brent 154Bernstein, Basil 65Bickerton, Derek 32–3, 140, 236bidialectalism 32, 34bilabial 13, 222bilingualism 31–2, 40, 73, 75, 86,

139, 150, 268binary antonymy 256binary feature 78binding 16, 104, 305biolinguistics 141bioprogram hypothesis 32–3,

60–61, 132, 140–1, 201, 222Black English 33–4, 86, 273blend 50, 167, 325Bloomfield, Leonard 14, 31, 105,

157, 177, 252, 277Boas, Franz 19, 157borrowing 34–5, 88, 100, 117,

136, 143, 159–60, 188, 202, 218,238, 248

bound morpheme 8, 177, 224branch, of language 153; of a

tree 307breath 213, 302Bresnan, Joan 149–50British Sign Language 260broad transcription 302

INDEX

352

Broca, Paul 20, 133, 160, 185Broca’s aphasia 20Broca’s area 20, 134, 160, 173, 185Brown, Roger 131Bruner, Jerome 140BSL see British Sign Languagebundle 78

calls 17–18, 197–8, 200canonical sentoid strategy 208caregiver speech 178caretaker speech see caregiverspeech

case 11, 35–7, 67, 75, 94, 102–3,106–8, 116, 203, 206, 311

case frame 94Catalan 63, 73, 113, 171–2, 196cataphora 16Chinese 10, 30, 73, 87, 109, 122,192, 202, 214, 271, 273, 294, 299,310, 327–8

Chomsky hierarchy 98Chomsky, Noam 28, 31, 47, 64, 79,93, 96–9, 103, 105, 119–20,126–7, 130–33, 140–41, 148,157–8, 171, 205, 209, 215, 217,220, 237, 251–2, 257, 277,289–90, 303–5, 308, 311–12

citation form 90, 324classical phoneme 215clause 37–8, 39–40, 48, 50, 53, 90,96, 175, 179, 218–19, 234, 241–2,245, 257, 278, 282–3, 288, 305–6

cleft 91click 12clipping 325clitic 38–9cliticization 39closed class 208cloze procedure 39, 84, 94Coates, Jennifer 258COBUILD dictionary 151Cockney 9–10code 39–40, 65code-switching 32, 39–40, 86codification 72–3, 105, 138, 230coding time 294–5cognitive grammar 41, 53, 129, 301cognitive linguistics 19, 28, 40–41,47, 49–50, 53–4, 66, 92, 94, 112,

128–9, 133, 147, 165–7, 169–70,174, 225, 229, 236–7, 250, 278,281, 298, 301, 308

cognitive metaphor 169cognitive poetics 50, 225, 281,

298, 309cognitive psychology 41, 237cognitive science 31, 41, 157, 167,

237, 281cognitive synonymy 236coherence 42, 65, 293, 296cohesion 4, 41–2, 76, 293, 296coindexing 17colligate 42, 49collocation 42colloquial speech 224, 232combining form 326comitative case 36comment 299communicative competence 43,

162, 224, 309community of practice 43–4, 268comparative linguistics 111comparative method 44, 46comparative reconstruction 44–6,

111, 116–17, 121, 235, 243–4, 292comparison 6competence 34, 39, 43, 47, 130–31,

148, 205, 209complement 24, 47–8, 90, 117, 329complement clause 38, 48, 283complementary distribution 80complementizer 48, 50completive aspect 26complex sentence 37compositional deixis 66compound sentence 37compounding 325computational linguistics 42,

48–9, 77, 149–51, 158, 185, 206,281, 303

computer-assisted languagelearning (CALL) 49

computerised tomography (CT) 161conceptual integration theory

49–50conclusive aspect 26concord see agreementconcordance 49, 149, 206conduction aphasia 20

INDEX

353

configurational language 53conjunction 48, 50, 58–9, 208conjuncts 58conlang 25connotation 34, 51, 67, 89, 165, 256consonant 9, 13, 45, 51–2, 121,123–4, 162–3, 188, 202, 213–14,216, 218, 222–3, 258, 270,283–4, 291–2, 298, 302, 311, 319,321–2, 327

constituent 52–3, 114, 118, 189,219, 307, 328

constituent structure 39, 52–3, 110,149, 206, 219–20, 245, 280, 287,300, 307–8, 329

constriction 162construction grammar 53–4contact see language contactcontent-process debate 28Context-Free Phrase-StructureGrammar (CF PSG) 98, 219

context-free rule 303continuity hypothesis 200continuous aspect 26contour 126contrastive distribution 80contrastive linguistics 21contrastive stress 276control 54–5, 242convergence 152conversation analysis 4–5, 55,76–7, 240

conversational implicature 56–7,58, 85, 228, 232–3, 246

converse pairs 256cooperative 56, 232, 258cooperative principle 57–8,228, 246

coordinate structure 50, 58–9coordinating conjunction 50copula 59, 284core grammar 104coreferential 17Cornish 63, 138, 173coronal 222corpus 49, 149, 151, 184corpus linguistics 42, 49, 77corpus stylistics 281counterpart 298covert prestige 231

creak 213creativity 198creole 32–4, 60–61, 132, 142–3,

152, 195, 222, 236, 261creole continuum 60creolization 60critical discourse analysis 61–2, 76,

89, 113, 187, 254, 293critical language awareness 61critical linguistics 61critical period hypothesis 28, 62,

131–2, 143crossing 86C-structure 149cut-off age 62

Dalgarno, George 24dative case 11, 35–6, 67, 103, 116,

203, 206daughter language 44, 46, 100, 111,

117, 141, 244Davidson, Donald 212dead language 62–3, 138,

230–1, 243declension 39decreolisation 60deep case see semantic roledeep dyslexia 82deep structure 63–4, 68–9, 171,

285, 304, 308deep-structure ambiguity 14defeasibility 56deficit hypothesis 65degree 6–7, 107, 174–5, 223degree-modifier 6deictic category 65–6, 107, 246deictic centre 66deixis 65–6demonstrative pronoun 234denotation 51, 66–7, 87, 165, 254–5dense social network 263dental 80, 115, 124, 214, 222dependency 67–8, 103derivation 9, 68–9, 118, 177–8, 207,

224, 247, 305, 325–6Descartes, Rene 24, 156descriptivism 69–70, 156, 184, 188,

203, 231, 248, 287, 301, 313design features 18, 23, 70, 78, 81,

198, 276

INDEX

354

determiner 11, 35, 38–9, 70–71, 94,96, 190–91, 204, 208, 220, 278,287–8, 308

determiner phrase 191developmental dyslexia 82deviance 92diachrony 71, 287diacritic 109, 123–4, 214dialect 4, 7, 32, 39–40, 72–3, 83,100, 116, 135, 140–1, 144, 192–3,211, 243, 273

dialect atlas 72dialect chain 136dialect continuum 193dialect geography 72, 265dialect map 72dialectic 247dialectology 72, 158, 216, 265difference hypothesis 65diffusion 72, 126, 135, 153, 263diglossia 32, 73–5, 265, 268digraph 109diphthong 15, 81direct complement 48direct object 35–6, 75, 108, 115–16,189, 251, 282

discontinuity hypothesis 200discourse 55, 61, 76, 242, 269discourse analysis 55, 76–7, 130,165, 243, 254, 280, 297, 303, 309

discourse world 297displacement 70, 77–8, 198, 276distinctive feature 78–9,183–4, 286

distribution 5, 79–80, 170, 207,215, 277, 289

ditransitive 306divalent 23double articulation 81D-structure 64, 104dual number 192duality of patterning 18, 70,80–81

Dutch 60, 88, 123, 142, 145, 171,182, 193–4, 196, 273

dying language 138Dyirbal 37, 97dysgraphia 82dyslexia 82dysphasia 19

ear training 216Ebonics 34egocentric deixis 66egressive 12elaborated code 65E-language 130elicitation techniques 82–3, 85elision 84ellipsis 36, 83–4, 95, 29embedded question 38empty category 17, 96enclitic 39ending 13, 24–5, 27, 29, 36, 46, 97,

108, 178, 222–3, 258, 274, 277,281, 328

Englishes 272entailment 57, 84–5, 232, 233, 297epithet 180equational sentence 59ergative case 35–7Esperanto 25essive case 36EST see Extended Standard Theoryethics 83, 85, 195ethics committee 85ethnicity 65, 86, 112ethnography of communication 19,

43, 86–7, 162, 187, 205, 224,240, 309

ethnolinguistics 86ethnomethodology 55etymology 87–8, 211–12euphemism 89, 256event time 295evidentiality 107evolution of language see origin

and evolution of languageExtended Standard Theory 305extension 67external sandhi 13–14extraction 67, 179extraposition 179

face 223Faiguet de Villeneuve, Joachim 24Fairclough, Norman 61falsetto 213family name 180Fant, Gunnar 79Fauconnier, Gilles 49–50, 166–7

INDEX

355

feature geometry 79feature specification 79feedback 5felicity condition 267feminine 11, 96–7feral children 62, 131Ferguson, Charles 73, 75, 265figure and ground 41, 92, 129, 301finger-spelling 166finite 10, 48, 89–91, 118, 279Finnish 25, 35–6, 110, 117, 146,159, 194, 210, 271, 273

first person 210–11Firth, J.R. 19, 156, 292–3Fishman, Joshua 86, 113, 265flap 52, 162–3Flemish 182, 194floor 309flout 58focus 91, 300Fodor, Jerry 212Foley, William, 94folk linguistics 91–2, 145folk taxonomy 19foregrounding 41, 92, 114,129, 301

forensic linguistics 93, 22formal grammars 93, 130, 188formal language 130formal logic 168–9formal semantics 66, 253, 255formal universal 311formulaic sequence 275fragment 257, 314frame 41, 70, 93–4, 128, 189–90,207, 250, 281, 297

free morpheme 177Frege, Gottlob 212French 3, 10, 15–16, 22–4, 30, 32,34–5, 60, 63, 75, 88–9, 96, 117,122, 124, 137, 152, 160, 168,171–2, 182, 197, 222, 235, 238–9,273, 277, 300, 323

fricative 51, 80, 124, 162–3, 214fronting 179F-structure 149Functional Grammar 293functional literacy 158functional sentence persepctive 102functional structure 149

functionalism 43, 94–5, 106, 251,290, 292–3, 296–7

functions of language 95

Gabelentz, Georg von der 156gap 17, 59, 67, 95–6, 283gapping 59garden-path strategy 208Gazdar, Gerald 220GB see Government-and-Binding

Theorygender 11, 43–4, 65, 96–7, 103,

107, 195, 258–9, 315general phonetics 216Generalized Phrase-Structure

Grammar (GPSG) 220generation 21, 47, 129, 138, 159,

172–3, 221, 230, 273generative grammar 93, 97–9,

106–7, 116, 127, 130, 150, 157,171, 198, 219–20, 248, 277–8,289, 300, 303,305

generative phonology 215, 217generative semantics 53, 304genetic hypothesis of language 28,

53, 99, 119–20, 132, 140–41, 201genetic relationship 99–100, 142,

211, 235, 310–11genitive case 35–6, 203genre 100–1, 125, 181, 298Georgian 128, 218German 10–11, 22, 34, 40, 67,

73–5, 88, 96–7, 103, 107, 109,112, 122, 124, 128, 142, 153, 163,171–2, 182, 193–4, 238

Germanic languages 116, 142, 235gerund 90Geschwind, Norman 20, 185gestural hypothesis 173given/new 91, 101–2, 299–300global aphasia 20global language 32, 122glotallic 12glottal 222, 264, 321glottis 12, 222, 269, 319Gopnik, Myrna 139government 11, 67–8, 102–3governmental concord 11Government-and-Binding

Theory 37, 64, 75, 99, 103–5,

INDEX

356

106, 127, 150, 171, 191, 219, 239,273, 277, 305, 308, 313

GPSG see GeneralizedPhrase-Structure Grammar

gradable antonymy 256grammar 53, 105–6grammatical category 26–7, 29, 35,65–6, 97, 106–7, 118, 174–5,191–2, 210–11, 294, 319

grammatical function seegrammatical relation

grammatical morpheme 177grammatical relation 24, 107–8, 149grammatical word–form 324grapheme 108–9, 123, 202graphology 326Greek 25, 35, 44–6, 74, 88, 110,116–17, 128, 138, 142, 152–3,326–7

Greenberg, Joseph 290, 310–12Grice, Paul 57–8, 212, 246ground, and figure 41, 92, 129, 301Guaranı 74

habitation name 197habitual aspect 26Halle, Morris 79, 215Halliday, Michael 42, 94, 200, 243,252, 293, 298, 307

Harris, Zelig 289Hawaiian 88, 159, 218, 221Hawkins, John 311–12head 11, 38, 71, 109–10, 175,189–90, 219, 288, 311, 328–9

Head-Driven Phrase-StructureGrammar (HD PSG) 220

headlining 91headword 110hedging 174hieroglyphs 109, 202, 328high-rising terminal 126historical linguistics 46, 71, 100,110–11, 121, 135–6, 142, 156–7,211–12, 243–4, 262, 292

Hittite 63, 116, 121Hockett, Charles 70, 78, 81,198, 276

Hogben, Lancelot 25home sign 261host 38

HPSG see Head–DrivenPhrase-Structure Grammar

Humboldt, Wilhelm von 154, 156,310–11

H–variety 74hydronym 197Hymes, Dell 43, 87hypercorrection 232hyponymy 256

iconicity 23, 111–12, 155, 266ideational function 293identity 33, 59, 86, 112–13, 114,

262–3, 275, 309ideology 89, 113–14, 130, 182,

247, 263ideophone 266idiom 39, 114–15, 242, 275I-language 130ill-formedness 323illiteracy 158illocution 267image-schema 92, 129, 229, 301imperfective aspect 26implicature 246implosive 12impressionistic phonetics 215inchoative aspect 26incorporation 224indefinite pronoun 234indicative mood 128indicators 115, 164, 231, 275indirect object 108, 115–16indirect question see embedded

questionIndo-European 46, 88, 110, 116–17,

121, 128, 142, 144, 153, 235infinitive 42, 90, 117–18, 230infix 9inflecting language 149, 299, 310inflection 9, 29, 35–6, 68–9, 118,

137–8, 149, 177–8, 202–3, 207, 299ingressive 12initialism 326injective 12innateness hypothesis 28, 33, 99,

119–20, 132–3, 140–41, 143,201, 312

insertion-sequence 5instrumental phonetics 215

INDEX

357

integrationalism 120, 130intension 255interface 27intermediate category 288internal reconstruction 46, 111,121, 244

internal sandhi 13international language 121–3, 152International Phonetic Alphabet 1,4, 123–5, 188, 213, 216, 302–3

interpersonal function 293interrogative pronoun 234intertextuality 125, 296, 298intonation 125–6, 134, 178, 234–5,283–4, 299, 314

intransitive 23, 35–6, 54, 75, 306intuition 83, 126–7Inuit 37, 88, 144IPA see International PhoneticAlphabet

Irish (Gaelic) 23, 30, 44–6, 88, 92irrealis 128isogloss 72isolating language 310Italian 34, 75, 87–9, 123, 188,235, 271

iterative aspect 26

Jakobson, Roman 79, 101, 131,278, 280

Japanese 30, 87, 137, 144, 202, 218,223, 229, 271, 277, 283, 299,310, 327–8

Javanese 87, 223Jespersen, Otto 25, 143Johnson, Dr Samuel 150Jones, Daniel 215–16Jones, William 110, 116juncture 284

Kabardian 311Kalaallisut 27Kanuri 299Kaplan, Ronald 149–50Katz, Jerrold 212Kay, Paul 154kinesics 187kinship terms 19, 128–9Klingon 25Kripke, Saul 212

Kristeva, Julia 125Kroeber, A.L. 157Kruszewski, Mikolaj 156, 215–16Kurdish 37

labial 80, 124, 222labio-velar 223Labov, William 21, 65, 72, 127,

158, 181–2, 196, 240–41, 264–5,315–16

LAD see language acquisitiondevice

Lakoff, George 41, 54, 170, 184, 237laminal 222landmark 41, 92, 129, 229, 301Langacker, Ronald 41, 53–4, 92,

129, 301langage 130, 140language 129–31language acquisition 28, 31, 54, 62,

119, 131–2, 133, 142–3, 158, 165,178, 237, 270, 309

language acquisition device 132–3language and ethnicity 86language and identity 113language areas 20–21, 133–4,

139–40, 148, 160–61, 185, 200language awareness 136language change 15–16, 21, 35, 44,

60, 71–2, 110–11, 134–6, 142,212, 232, 241, 244, 249, 263,271, 316

language conflict 32language contact 34–5, 46, 60, 100,

135, 136–7, 138, 153, 159–60,238, 265

language death 63, 137–8, 173,194–5

language disability 21–2, 28, 99,134, 139–40, 152

language faculty 20, 27–8, 47, 99,130–31, 139, 140–1, 143, 201

language family 46, 100, 110–11,116–17, 135–6, 141–2, 153, 235,291–2

language in use 296language instinct 31, 62, 99, 132–3,

142–3, 261language myths 92, 136, 143–5language pathology 270

INDEX

358

language planning 22, 100, 146–7,260, 263

language processing 134, 141, 147,160, 208–9, 237, 246, 297

language shift 138language testing 39langue 47, 120, 130, 147–8, 205, 209laryngeal 121laryngeal hypothesis 121larynx 12–13, 213, 319–20late closure, principle of 208lateral 51, 163lateral affricate 163lateral approximant 163lateral fricative 163lateralization 148, 185Latin 10, 15–16, 24–5, 35, 44–5, 63,71, 75, 88, 96, 107, 110, 116–17,122, 150, 164, 203–4, 212, 231,300, 310, 326

Leibniz, Gottfried von 24lemma 68, 90, 118, 148–9, 324–5Lenneberg, Eric 62, 131letter 17, 37, 82, 108–9, 123, 166,201–2, 303, 326–7

Letzebuergesch 194levels in intonation 126Levi-Strauss, Claude 278lexeme see lexical itemlexical ambiguity 14lexical aspect 27lexical category see part of speechlexical cohesion 42lexical entry 149lexical head 110, 328lexical item 16, 220, 224, 324lexical morpheme 176lexical semantics 253Lexical-Functional Grammar 106,149–50

lexicalisation 34, 89, 237lexicography 22, 49, 150–51lexicology 243, 325lexicon 27, 149, 151–2, 220LFG see Lexical-FunctionalGrammar

Linear B 327lingua franca 152, 195, 221linguicide 138linguistic anthropology 19

linguistic area 137, 153linguistic competence 43linguistic engineering 146linguistic relativity 19, 154–5linguistic sign 23, 155, 252–4, 286linguistics 156–8lip-reading 109lisping 270literacy 158–9, 166, 199–200literary linguistics 281loan word 35, 88, 136–7, 159–60local dependency 67localization 160–1, 185Locke, John 156locution 267logical form (LF) 104loglang 25logographic writing system 326–7logonomic rules 161–2long passive 320Lucy, John 154L-variety 74

machine translation 49magnetic resonance imaging (MRI)

20, 161main clause 37main stress 276Malagasy 30, 223Malinowski, Bronislaw 19, 156manner of articulation 9, 13,

162–3, 223mapping 50, 169–70, 308Margi 299marked form 148, 163–4, 259markedness 163–4, 199, 320markers 13, 29, 66, 115, 164, 223,

275, 283, 303Martinet, Andre 81masculine 11, 96–7, 203matched guise test 83Mathesius, Vilem 102matrix clause 78maxims of conversation 58, 246meaning 165median consonant 163medium 70, 134, 165–6, 260, 266–7mental lexicon 152mental model 17, 41mental space 49, 166–7

INDEX

359

mentalism 31, 41, 157, 167meronymy 256mesoclitic 39mesolect 60metafunction 293metalanguage 167–9, 188metaphor 19, 41, 49–50, 89, 115,129, 165, 167, 169–70, 187, 256,298, 301, 308–9

metaphorical code-switching 40metrical phonology 217Milroy, Jim 136, 158, 231, 239, 241,249, 274, 318–19

Milroy, Lesley 158, 196, 231,239–40, 263, 265, 274, 318–19

minimal pair 80, 83, 170, 186,215, 217

minimalist program(me) 105, 171,199, 305

minority language 32, 114, 137–8,171–3, 182, 197

mirror neurons 173–4modal auxiliary 29, 35, 174modality 29, 174, 176, 295modalization 112, 174, 298model 27, 98, 179, 203, 206model-theoretic semantics 253modification 175modifier 6–7, 38, 107, 174–5, 219,234, 261, 311

modularity 141module 27, 104, 220, 249mondegreen 14–15monovalent verb 23Montague, Richard 253mood 107, 117, 128, 174, 175–6, 274moribund language 138morpheme 8–9, 13, 52, 80, 151, 164,176–7, 178, 183–4, 224, 243, 247,278, 288, 294, 310, 323–4, 325–6

morphemographic writingsystem 326

morphology 27, 37, 68–9, 105–6,118, 145, 152, 164, 177, 178, 199,203, 235, 243, 247, 274, 289,310–11, 323, 326

morphosyntax 203mother language 141motherese 178–9movement 179–80, 242

multilingual 31multiplex social network 263Murray, James 150

name 19, 22, 72, 135–6, 150, 155,159, 170, 180–81, 188, 197, 211,223, 245, 260, 267, 303, 320, 327

narrative 19, 66, 83, 112, 181–2, 227narrative aspect 26narratology 181narrow transcription 302nasal 51, 79, 123–4, 162, 205, 250,

269–70, 311, 319national language 32, 75, 114,

146–7, 171–2, 182, 195, 197, 273natural class 78–9, 182–4natural language 83, 130, 184,

206, 260natural-language processing 25, 49,

131, 184–5, 206, 222, 261nature-nurture debate 28Navaho 97, 172, 223N-bar 190, 220, 278, 288, 308, 329near-RP 3near-universal 311negative face 223negative raising 242Neogrammarians 111neuro-linguistic programming 185neurolinguistics 20–21, 41, 134,

140, 158, 161, 185, 237neuter 96neutralization 186, 215, 217Nicaraguan Sign Language 261NLP see natural-language

processingnode 307–8node label 307–8nominal group see N-barnominalization 89–90, 108,

186–7, 282nominative case 35–6, 75, 102,

203, 206non-configurational language 53non-finite 48, 89–91, 117non-linear phonology 217non-modal auxiliary 29non-specific indefinite 246non-verbal communication 86–7,

187, 205

INDEX

360

notational convention 93, 188, 248noun 6, 8–9, 11, 13, 35–6, 38, 71,80, 96–7, 106–7, 110, 118, 139,178, 180, 183, 186–7, 188–9,191–2, 203–8, 233, 245, 248, 276,278, 282, 287–8, 311, 329

noun phrase 23, 35, 37, 48, 50,54, 58–9, 64, 67, 70–71, 75,80, 96, 104, 106–8, 110, 116,175, 180, 186–7, 188, 189–91,218–20, 228–9, 234, 243–5,178–9, 282, 284, 287–9, 300,305–6, 317, 329

noun-complement clause 48, 283NP see noun phrasenull-element 17number 6, 10–11, 106–7, 191–2,207, 211

number of languages 192–5

object 11, 23, 30, 54, 67, 75, 96,102–3, 149, 179, 189–90, 218,229, 242, 285, 294, 306

object language 168object-complement 48object-control 54objective case 102–3object-to-subject raising 242oblique object 108observer’s paradox 83, 85, 127,195–6

obstruent 51, 162obviative 211oesophagic egressive 13official language 114, 122, 146–7,171, 182, 196–7, 273, 328

Ogden, C.K. 25Okrand, Marc 25Omaha system 129one-place predicate 228onomastics 73, 88, 181, 197,211–12

onomatopoeia 112open class 208open-endedness 18, 70, 78, 81,197–8, 276–

optative mood 128optimality theory 198–9oracy 159, 199–200orality 199

origin and evolution of language 18,46, 87, 95, 100, 110, 116, 141, 145,194–5, 179, 200–1, 236, 310

oronym 197orthographic word 324–5orthography 109, 201–2, 238,

272, 328overlexicalization 89, 144, 262Oxford English Dictionary 151

palatal 222, 270, 319palato-alveolar 222paradigm 202–3, 204paradigmatic relation 203, 204,

289, 291paralanguage 12–13, 87, 187, 204–5parameter 104parameter-setting model 133paraphrase 85parole 47, 120, 130, 148, 205,

209, 267parser 93, 184, 206parsing 108, 205–6part of speech 5–6, 8, 29, 48, 50,

70–71, 152, 184, 188–9, 204, 206–8,228, 233, 282, 287–8, 316, 325

participant 269participant observation 83, 240participant role see semantic rolepassive participle 90passive vocabulary 151passive voice 320passive-with-agent 320past participle 90patronym 180paucal number 192peak of sonority 322Peano, Giuseppe 25Peirce, C.S. 253perceptual phonetics 215perceptual strategy 147, 208–9perfect aspect 26perfective aspect 26performance 47, 130–31, 148, 196,

205, 209, 267performative 209–10, 267perlocution 267Persian 34, 88, 194person 10–11, 66, 90, 107, 117, 192,

210–11, 311

INDEX

361

personal name 178, 197personal pronoun 102, 211, 234, 291PET scanner 161Peters-Ritchie results 305philology 197, 211–12philonymy 256philosophy of language 169,212–13, 225, 280

phonaestheme 266phonaesthesia 112, 266phonation type 13, 213, 321phoneme 78–81, 109, 124, 137, 155,170, 183, 186, 202, 213–15,217–18, 243, 250, 270–71, 278,286, 288, 294, 302–3, 321, 327

phonemic transcription 302–3phonetic transcription 269, 302–3phonetics 10, 14, 123–4, 163,215–16, 217, 234, 250, 285, 303,319, 321

phonological word 324phonology 4, 10, 14, 22, 39, 44, 46,69, 78–80, 117, 121, 133, 164,171, 186, 188, 199, 215, 216–17,218, 234–5, 243, 248, 250, 284–6,289, 294, 302–3, 322, 324–5

phonotactics 14, 80–81, 170,217–18

phrasal category 218, 287phrase 24, 38, 40, 53, 58–9, 89, 96,109–10, 114, 164, 175, 179,218–19, 243–5, 278, 287–88, 294,315, 323–4, 326

phrase marker see treephrase-structure grammar 26, 53,98–9, 106, 110, 150, 206, 219–21

phrase-structure rule 220physiological phonetics 215Piaget, Jean 119, 140pidgin 33–4, 60–61, 132, 142, 152,195, 221–2, 236

Pinker, Steven 143Piraha 214pitch accent 283, 299pitch language 234place name 180, 197place of articulation 162–3, 222–3plesionymy 256plosive 9–10, 13, 51, 162, 214plural 188, 191

politeness 10, 43, 144, 223–4, 227Pollard, Carl 220–21polysynthesis 224Port-Royal grammarians 105positive face 223positron emission tomography

(PET) 161possessive pronoun 71possible worlds theory 167, 225,

250, 298, 323post-Bloomfieldians see also

American structuralists 157post-modifier 6postposition 229poverty of the stimulus 120, 132power 43, 114, 123, 138, 196,

225–6, 227, 258–9, 273, 309pragmatics 54, 57–8, 87, 120, 130,

158, 184, 226–8, 232, 246, 248–9,251, 253, 257, 265, 267, 280

Prague School 101predeterminer 71predicate 37, 108, 228, 284, 289,

294, 317–18predicate logic 228predicate nominal 108prefix 6, 9, 45, 68, 176–7, 207, 233,

278, 291preliteracy 159pre-modifier 6prepalatal 222preposition 39, 67, 102–3, 108,

110, 129, 168, 179, 190, 206–8,218, 228–9, 231, 248, 287, 301,311, 329

prepositional phrase 24, 58, 110,137, 188, 190, 218, 229, 244, 329

prescriptivism 8, 69–70, 92, 100,136, 145, 229–31, 239, 248, 260,262, 272, 301

prestige 60, 72–4, 135, 137–8, 156,159–60, 172–3, 221, 225–6,231–2, 263, 315

presupposition 57, 85, 232–3, 297primary auxiliary 29primary medium 70, 165–6primary stress 276primitive languages 143–4Prince, Alan 198principle of late closure 208

INDEX

362

principles of grammar 104–5, 130,157, 312–13

Principles-and-Parameters see alsoGovernment-and-BindingTheory 103

proclitic 38productivity 198, 233, 275progressive aspect 26pronoun 10, 17, 35–6, 71, 97,102–3, 106–7, 160, 191, 208,210–11, 233–4, 259, 282, 291

proper name 245prosodic phonology 234–5prosody 126, 234–5, 284Proto-Indo-European 46, 87, 116,121, 235

proto-ndash;language 100, 111, 235protolanguage hypothesis 18, 201,235–6

prototype 181, 236–7, 255, 282, 306proximity principle 112psycholinguistics 17, 19, 39, 92,120, 147, 150, 155, 158, 165, 204,208–9, 237, 289

pulmonic 12punctual aspect 26punctuation 109, 201–2, 238, 328purism 137, 231, 238–9purpose complement 48

qualitative approach 95, 196,239–40

quantifier 71quantitative approach 21, 196,240–41, 249, 316

quantity principle 112Quine, Willard van Orman 212Quirk, Randolph 301

radiality 236raising 55, 179–80, 241–2Rampton, Ben 86rank scale 242–3, 282real English 318real time 21reanalysis 325–6Received Pronunciation 231reciprocal pronoun 234reconstruction 44–6, 74, 87, 111,116–17, 121, 138, 235, 243–4

recursion 81, 98, 244–5, 280, 288,290, 329

reference 41, 66–7, 104, 165, 169,212, 225, 245–6, 282, 290

reference time 295referential indices 17referring expression 245reflexive pronoun 234regional dialect 32, 44, 72, 100, 135,

141, 193, 235, 261, 265, 291, 318relational deixis 66relative clause 38, 175, 234–5, 283relative pronoun 234relevance 57–8, 228, 246Relevance Theory 246relexification 61resonant 51REST see Revised Extended

Standard Theoryrestricted code 65retroflex 222reverse click 12Revised Extended Standard

Theory 304–5rheme 102rhetoric 169, 238, 246–7, 256,

308–9rhoticity 315right-node raising 59role 65–6, 97, 269, see also

semantic roleRole-and-Reference Grammar

(RRG) 94root 9, 117, 121, 244, 247, 274RP, see Received Pronunciationrule 188, 248, 269, 322–4Russell, Bertrand 212, 232Russian 10, 75, 88, 107, 117, 122,

142, 146, 277, 310, 328

Sacks, Harvey 55, 309Sadock, Jerrold 27Sag, Ivan 220Sampson, Geoffrey 99sandhi 13Sapir, Edward 19, 154, 157, 215,

252, 277, 310Sapir-Whorf hypothesis see also

linguistic relativity hypothesis154–5, 185

INDEX

363

Saussure, Ferdinand de 47, 71,120–21, 130, 140, 148, 155–6,169, 204–5, 209, 248, 252, 254,265, 277–8, 287–9

Saussurean paradox 71, 111, 241,248–9, 265, 316

S-bar 288Scale-and-CategoryGrammar 293

schema 41–2, 94, 173, 225,249–50, 298

Schleyer, Johann Martin 24Schuchardt, Hugo 60scripts 41Searle, John 267second person 210secondary medium 166secondary stress 276segment 250, 270, 284, 286segmental phonology 250segregationalism 120selection restriction 165,250–51, 252

semantic feature 128semantic role 210, 228, 251–2semantics 16, 23, 27, 42, 53–4, 66,94, 158, 165, 212, 226–8, 232,236–7, 243, 249, 251, 252–3,255–6, 280, 304, 323

semi-modals 29semiology see semioticssemiotics 76, 155, 253–4, 278sense 254–5sense relation 253, 255–6sense relation network 83sentence 37–8, 256–7, 283,289–90, 314

sentence adverb 7–8sentence stress 276sequential order principle 112setting 104, 269sex differences in language 19,96–7, 258–9, 265, 309, 316

sexist language 146–7, 259–60sharing 54short passive 320shorthand 166Shweder, Richard 154side-sequence 5sign see linguistic sign

sign language 20, 70, 81, 132, 134,140, 142–3, 166, 260–61, 266

significant 155signifie 155simple sentence 37singular 6, 10–11, 13, 29, 71, 90,

106, 118, 188, 191, 207, 211,259, 324

situation semantics 253situational code–switching 40Skinner, B.F. 31, 131skip-connecting 5slang 261–2, 313SLI see Specific Language

Impairmentsluicing 84Smolensky, Paul 198sobriquet 180social dialect 72social history of language 114,

262–3social network 44, 263, 268social semiotics 254social stratification of language 60,

113, 241, 264, 265, 268, 316sociolect 72sociolinguistic competence 43sociolinguistics 19, 21–2, 34–5, 39,

43–4, 47, 72, 76, 82–3, 85–6, 97,111, 115, 127, 130, 135, 157–8,164–5, 170–71, 178, 195–6, 165,170–71, 178, 195–6, 199, 223,227, 231–2, 239, 241, 263, 264–5,272, 275, 303, 309, 314–16, 318

sonorant 162sound symbolism 112, 266source domain 50Spanish 10–11, 22, 52, 60, 63, 75,

88, 112–13, 117, 124, 142, 152,162–3, 171, 187, 196, 212, 214,235, 247, 274, 277, 321

specific indefinite 245Specific Language Impairment

28, 139specifier 175, 329speech 266–7speech act 210, 228, 266, 267speech community 32, 43–4, 73,

85–7, 91, 130, 138, 151, 263,265–6, 268, 275, 315

INDEX

364

speech defect 20, 139speech error 147, 209, 267speech event 268–9speech planning 147speech situation 234speech sound 12–13, 51–2, 78,80–81, 123–4, 153, 215–16,248–50, 269–70, 277, 302, 320–22

speech synthesis 49speech therapy 270spelling 4, 13, 80, 105, 109, 150–1,177, 201–2, 231, 269–70, 303,313, 327–8

Sperber, Dan 58, 246split infinitive 230Sprachbund see also linguistic

area 153S-structure 104stadialism 144stammering 82, 139, 270standard English 34, 199–200,225–6, 262, 272–3

standard language 32, 65, 72–3,159, 182, 195, 226, 229–30,272–4, 318

Standard Theory 304–5standardization 72, 105, 201stem 15–16, 117, 176–8, 274stereotype 115, 164, 255, 275stigma 231stimulus-bound 275stimulus-freedom 18, 70, 78, 198,275–6

story grammar 181strategic competence 43Strawson, Peter 212, 232stress 15–16, 38, 91, 234–5, 276–7,283–4

structural ambiguity 14structuralism 40, 79, 94, 217, 249,254, 277–8, 279, 281, 290–91

structure 278–9structure-dependence 279–80, 290stuttering 139, 270stylistics 22, 49, 92, 225, 247, 250,280–1, 293, 298, 309

subcategorization 29, 67–8, 103,183–4, 208, 281–2, 288, 317

subject 10–11, 23, 30, 35–7, 48,54–5, 59, 67, 75, 96, 107–8, 115,

137, 149, 186–7, 189–90, 228,241–2, 279, 282, 284–5, 294, 306,317, 319–20

subject-complement 48subject-control 54subject-raising 241subject-to-object raising 242subjunctive mood 128subordinate clause 37–8, 283,

241, 288subordinating conjunction 50subordination 37–8, 282–3subordinator 50substantive universal 311sub-world 297suffix 6, 8–9, 68, 90, 176–7, 207,

233, 259, 274, 278superfix 9superordinacy 256suprasegmental 126, 235, 250, 277,

283–4, 299surface structure 37, 64, 68–9, 104,

171, 284–5, 304surface-structure ambiguity 14surname 180, 197Swahili 10, 30, 129, 152, 310Sweet, Henry 216syllabary 202, 327–8syllable 81, 186, 202, 218, 270,

276–7, 279, 283–4, 285–6, 298–9,322, 327

symbol 326–7symbolic system 155, 254, 286–7synchrony 71–2, 287synonym 89, 256, 294syntactic category 53, 58, 190, 218,

244, 280, 282, 287–8, 290, 307–8,317–18

syntagm (syntagma) 279, 288–9syntagmatic relation 42, 204, 279,

288–9syntax 27, 38–9, 53–4, 59, 76,

84, 105–6, 108, 115, 171,188, 203, 220, 229, 243,245, 257, 270, 279–80, 282,289–90, 323

system 290–1systematic correspondence 45–6,

291–2systematic phoneme 215

INDEX

365

Systemic Linguistics 42, 94–5, 102,113, 174, 187, 242–3, 292–3,296–7, 306

tag question 258Tagalog 9, 37tagging 29tagmemics 294tag-question 258Tamil 34, 73, 268Tannen, Deborah 258–9tap 52, 162target domain 50Tarski, Alfred 212telic 27temporal case 36tense 6–7, 10, 16, 26–7, 29, 66,89–91, 107, 117, 177, 192, 211,294–5, 274, 281, 316–7, 326

tertiary medium 166text 42, 61, 76, 94, 101, 125, 181,242, 293, 295–6, 297

text linguistics 76–7, 243, 296–7, 298text world 297text world theory 147, 225, 250,296–7

textual deixis 66textual function 293textuality (texture) 296–7, 298, 309TG (TGG) see transformationalgrammar

thematic role see semantic roletheme 102, 251–2theoretical linguistics 22, 28, 83,130, 184, 188, 280

theta-role see semantic rolethird person 10–11, 90, 101, 210–11Thomas, John 60Tibetan 37Tolkien, J.R.R. 25tone 162, 205, 214, 223, 284, 298–9tone language 234–5, 284,298–9, 310

topic 91, 102, 179, 180, 299–300topicalization 179, 300toponym 197, 180tough-movement 242traditional grammar 37, 48, 50,69, 106, 108, 117, 183, 188,300–301, 317

trajector 41, 92, 129, 229, 301transcription 77, 215–6, 269,

301–3transformation 54, 69transformational generative

grammar see transformationalgrammar

transformational grammar 64, 68,99, 104–6, 157, 171, 217, 237,285, 303–5, 329

transitivity 24, 36–7, 75,305–7, 317

tree (diagram, tree structure) 14, 38,53, 59, 69, 149, 219–20, 284–5,307–8, 329

trill 52, 124, 163trivalent verb 23tropes 169, 308–9Trubetzkoy, Nikolai 79, 186Trudgill, Peter 264truth-conditional semantics 253tune 126Turing test 206Turkish 30, 144, 152–3, 159, 210,

239, 310, 328turn-taking 5, 55, 162, 258–9, 309two-place predicate 228typology 30, 290, 309–10, 312

unbounded dependency 67underlexicalization 144underlying form 69universal 70, 107, 118–9, 125, 129,

145, 144, 157, 206, 210, 234,311–12, 316

universal grammar 47, 104–6,119–20, 130, 140, 157, 312–3

universal translator 23unmarked form 117, 163–4Uralic 110urban dialectology 72usage 44, 49, 72, 76, 86, 126–8, 151,

224, 229–30, 239, 248, 313use-mention distinction 168utterance 257, 314

valency 23Van Name, Addison 60Van Valin, Robert 94variable 115, 241, 314–15, 316

INDEX

366

variation 3, 9, 72, 111, 135, 192,194, 229, 240–41, 249, 264–5,276, 315–16

velar 123, 222–3velaric 12velum 12, 51, 79, 162, 222, 269, 319verb 316–17verb phrase 29, 54, 110, 118, 175,186, 188, 219–20, 228, 241, 279,284, 287, 308, 317–18

verb–complement clause 48, 283vernacular 85, 113, 115, 179, 262,273, 318–19

Vietnamese 118, 153, 224, 310viseme 109visibility 107vocal folds (vocal cords) 213, 319,320–21

vocal tract 12, 51, 81, 162, 165,222, 319, 321

vocal-auditory channel 70vocative case 36voice 29, 107, 149, 319–20voice recognition 49voiceless implosive 12, 183voicing 79, 163, 213, 319, 320–21Volapk 24–5vowel 51–2, 124, 162, 216, 270,321–2

VP see verb phraseVP-deletion 84

Webster, Noah 150Weinreich, Uriel 249, 265well-formedness 27, 39, 76, 83, 251,257, 322–4

Welsh 22–3, 25, 30, 40, 88, 113,123, 138, 142, 163, 172, 180

Wernicke, Carl 20, 133, 160, 185Wernicke’s aphasia 20Wernicke’s area 20, 134, 160, 185whisper 213WH-movement 179Whorf, Benjamin Lee 154–5Wilkins, John 24Williams syndrome 28, 139Wilson, Deirdre 58, 246word 324–5word blindness 82word class see also part of

speech 106word-formation 69, 151–2, 177–8,

224, 233, 291, 325–6word-stress 276world language 32, 122world switch 297–8worlds 225, 297–8, 323writing system 4, 25, 108–9, 159,

166, 202, 238, 271–2, 301,326–8

W-star language 53

X–bar 288, 328–9xenonymy 256

Yimas 22

Zamenhof, L.L. 25zero determiner 71zero-element 17

!Xu 214

INDEX

367

Fifty Key Contemporary Thinkers

John Lechte

Fifty Key Contemporary Thinkers surveys the most important figures

who have influenced post-war thought. The reader is guided through

structuralism, semiotics, post-Marxism and Annales history, on to

modernity and postmodernity. With its comprehensive biographical

and bibliographical information, this book provides a vital refer-

ence work for all those who want to understand the intellectualhistory of the last fifty years.

ISBN 978-0-415-07408-7

Available at all good bookshops

For ordering and further information please visit

www.routledge.com

Psycholinguistics: The Key Concepts

John Field

Psycholinguistics: The Key Concepts is an authoritative, wide-ranging

and up-to-date A to Z guide to this important field. Cross-referenced,

with suggestions for further reading and a full index, the book is an

highly accessible introduction to the main terms and concepts in

psycholinguistics. Psycholinguistics offers over 170 entries covering

the key areas:

� Psychological processes

� First language acquisition

� The nature of language

� Brain and language

� Language disorders

This comprehensive guide is an essential resource for all students ofEnglish language, linguistics and psychology.

ISBN 978-0-415-25891-3

Available at all good bookshops

For ordering and further information please visit

www.routledge.com

Language and disguistic

Education

Transcript of Language and disguistic